Model save
Browse files- README.md +58 -0
- all_results.json +7 -0
- train_results.json +7 -0
- trainer_state.json +1282 -0
README.md
ADDED
|
@@ -0,0 +1,58 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
base_model: Qwen/Qwen2.5-7B
|
| 3 |
+
library_name: transformers
|
| 4 |
+
model_name: Qwen2.5-7B-SFT
|
| 5 |
+
tags:
|
| 6 |
+
- generated_from_trainer
|
| 7 |
+
- trl
|
| 8 |
+
- sft
|
| 9 |
+
licence: license
|
| 10 |
+
---
|
| 11 |
+
|
| 12 |
+
# Model Card for Qwen2.5-7B-SFT
|
| 13 |
+
|
| 14 |
+
This model is a fine-tuned version of [Qwen/Qwen2.5-7B](https://huggingface.co/Qwen/Qwen2.5-7B).
|
| 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="YYT-t/Qwen2.5-7B-SFT", 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/yifeizuo2029-northwestern-university/BRiTER/runs/ulquktyn)
|
| 31 |
+
|
| 32 |
+
|
| 33 |
+
This model was trained with SFT.
|
| 34 |
+
|
| 35 |
+
### Framework versions
|
| 36 |
+
|
| 37 |
+
- TRL: 0.15.2
|
| 38 |
+
- Transformers: 4.49.0
|
| 39 |
+
- Pytorch: 2.5.1
|
| 40 |
+
- Datasets: 3.3.2
|
| 41 |
+
- Tokenizers: 0.21.0
|
| 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 Gallouédec},
|
| 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 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"total_flos": 1.846643377789993e+17,
|
| 3 |
+
"train_loss": 0.30200490220900506,
|
| 4 |
+
"train_runtime": 1464.9032,
|
| 5 |
+
"train_samples_per_second": 6.741,
|
| 6 |
+
"train_steps_per_second": 0.106
|
| 7 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"total_flos": 1.846643377789993e+17,
|
| 3 |
+
"train_loss": 0.30200490220900506,
|
| 4 |
+
"train_runtime": 1464.9032,
|
| 5 |
+
"train_samples_per_second": 6.741,
|
| 6 |
+
"train_steps_per_second": 0.106
|
| 7 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1282 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 5.0,
|
| 5 |
+
"eval_steps": 100,
|
| 6 |
+
"global_step": 155,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.03225806451612903,
|
| 13 |
+
"grad_norm": 0.16533233225345612,
|
| 14 |
+
"learning_rate": 1.25e-08,
|
| 15 |
+
"loss": 0.3288,
|
| 16 |
+
"mean_token_accuracy": 0.9142395853996277,
|
| 17 |
+
"step": 1
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.06451612903225806,
|
| 21 |
+
"grad_norm": 0.16388170421123505,
|
| 22 |
+
"learning_rate": 2.5e-08,
|
| 23 |
+
"loss": 0.3143,
|
| 24 |
+
"mean_token_accuracy": 0.919180154800415,
|
| 25 |
+
"step": 2
|
| 26 |
+
},
|
| 27 |
+
{
|
| 28 |
+
"epoch": 0.0967741935483871,
|
| 29 |
+
"grad_norm": 0.16326533257961273,
|
| 30 |
+
"learning_rate": 3.75e-08,
|
| 31 |
+
"loss": 0.3239,
|
| 32 |
+
"mean_token_accuracy": 0.9184285998344421,
|
| 33 |
+
"step": 3
|
| 34 |
+
},
|
| 35 |
+
{
|
| 36 |
+
"epoch": 0.12903225806451613,
|
| 37 |
+
"grad_norm": 0.14039699733257294,
|
| 38 |
+
"learning_rate": 5e-08,
|
| 39 |
+
"loss": 0.2934,
|
| 40 |
+
"mean_token_accuracy": 0.9237579107284546,
|
| 41 |
+
"step": 4
|
| 42 |
+
},
|
| 43 |
+
{
|
| 44 |
+
"epoch": 0.16129032258064516,
|
| 45 |
+
"grad_norm": 0.17062482237815857,
|
| 46 |
+
"learning_rate": 6.25e-08,
|
| 47 |
+
"loss": 0.3162,
|
| 48 |
+
"mean_token_accuracy": 0.9196513891220093,
|
| 49 |
+
"step": 5
|
| 50 |
+
},
|
| 51 |
+
{
|
| 52 |
+
"epoch": 0.1935483870967742,
|
| 53 |
+
"grad_norm": 0.16192646324634552,
|
| 54 |
+
"learning_rate": 7.5e-08,
|
| 55 |
+
"loss": 0.3403,
|
| 56 |
+
"mean_token_accuracy": 0.9099262356758118,
|
| 57 |
+
"step": 6
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 0.22580645161290322,
|
| 61 |
+
"grad_norm": 0.14392420649528503,
|
| 62 |
+
"learning_rate": 8.75e-08,
|
| 63 |
+
"loss": 0.2809,
|
| 64 |
+
"mean_token_accuracy": 0.926060140132904,
|
| 65 |
+
"step": 7
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.25806451612903225,
|
| 69 |
+
"grad_norm": 0.16454868018627167,
|
| 70 |
+
"learning_rate": 1e-07,
|
| 71 |
+
"loss": 0.3227,
|
| 72 |
+
"mean_token_accuracy": 0.9182868003845215,
|
| 73 |
+
"step": 8
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.2903225806451613,
|
| 77 |
+
"grad_norm": 0.13043281435966492,
|
| 78 |
+
"learning_rate": 1.125e-07,
|
| 79 |
+
"loss": 0.2912,
|
| 80 |
+
"mean_token_accuracy": 0.9230725765228271,
|
| 81 |
+
"step": 9
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 0.3225806451612903,
|
| 85 |
+
"grad_norm": 0.13768798112869263,
|
| 86 |
+
"learning_rate": 1.25e-07,
|
| 87 |
+
"loss": 0.2853,
|
| 88 |
+
"mean_token_accuracy": 0.9282448291778564,
|
| 89 |
+
"step": 10
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"epoch": 0.3548387096774194,
|
| 93 |
+
"grad_norm": 0.1535075604915619,
|
| 94 |
+
"learning_rate": 1.375e-07,
|
| 95 |
+
"loss": 0.289,
|
| 96 |
+
"mean_token_accuracy": 0.9233593940734863,
|
| 97 |
+
"step": 11
|
| 98 |
+
},
|
| 99 |
+
{
|
| 100 |
+
"epoch": 0.3870967741935484,
|
| 101 |
+
"grad_norm": 0.174004465341568,
|
| 102 |
+
"learning_rate": 1.5e-07,
|
| 103 |
+
"loss": 0.3091,
|
| 104 |
+
"mean_token_accuracy": 0.9212210774421692,
|
| 105 |
+
"step": 12
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 0.41935483870967744,
|
| 109 |
+
"grad_norm": 1713721835520.0,
|
| 110 |
+
"learning_rate": 1.6249999999999998e-07,
|
| 111 |
+
"loss": 0.2708,
|
| 112 |
+
"mean_token_accuracy": 0.9274526238441467,
|
| 113 |
+
"step": 13
|
| 114 |
+
},
|
| 115 |
+
{
|
| 116 |
+
"epoch": 0.45161290322580644,
|
| 117 |
+
"grad_norm": 0.13813622295856476,
|
| 118 |
+
"learning_rate": 1.75e-07,
|
| 119 |
+
"loss": 0.2719,
|
| 120 |
+
"mean_token_accuracy": 0.9287762641906738,
|
| 121 |
+
"step": 14
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.4838709677419355,
|
| 125 |
+
"grad_norm": 0.15647657215595245,
|
| 126 |
+
"learning_rate": 1.875e-07,
|
| 127 |
+
"loss": 0.3118,
|
| 128 |
+
"mean_token_accuracy": 0.9167627692222595,
|
| 129 |
+
"step": 15
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.5161290322580645,
|
| 133 |
+
"grad_norm": 0.1404537558555603,
|
| 134 |
+
"learning_rate": 2e-07,
|
| 135 |
+
"loss": 0.2887,
|
| 136 |
+
"mean_token_accuracy": 0.9297909140586853,
|
| 137 |
+
"step": 16
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 0.5483870967741935,
|
| 141 |
+
"grad_norm": 0.16780979931354523,
|
| 142 |
+
"learning_rate": 1.9997445995478116e-07,
|
| 143 |
+
"loss": 0.3192,
|
| 144 |
+
"mean_token_accuracy": 0.9215397834777832,
|
| 145 |
+
"step": 17
|
| 146 |
+
},
|
| 147 |
+
{
|
| 148 |
+
"epoch": 0.5806451612903226,
|
| 149 |
+
"grad_norm": 0.16038931906223297,
|
| 150 |
+
"learning_rate": 1.998978528650029e-07,
|
| 151 |
+
"loss": 0.3057,
|
| 152 |
+
"mean_token_accuracy": 0.9223698377609253,
|
| 153 |
+
"step": 18
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 0.6129032258064516,
|
| 157 |
+
"grad_norm": 0.15409325063228607,
|
| 158 |
+
"learning_rate": 1.9977021786163597e-07,
|
| 159 |
+
"loss": 0.2906,
|
| 160 |
+
"mean_token_accuracy": 0.9280111193656921,
|
| 161 |
+
"step": 19
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 0.6451612903225806,
|
| 165 |
+
"grad_norm": 0.13709405064582825,
|
| 166 |
+
"learning_rate": 1.995916201407555e-07,
|
| 167 |
+
"loss": 0.2963,
|
| 168 |
+
"mean_token_accuracy": 0.9227046966552734,
|
| 169 |
+
"step": 20
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 0.6774193548387096,
|
| 173 |
+
"grad_norm": 0.17100510001182556,
|
| 174 |
+
"learning_rate": 1.9936215093023882e-07,
|
| 175 |
+
"loss": 0.3308,
|
| 176 |
+
"mean_token_accuracy": 0.9156600832939148,
|
| 177 |
+
"step": 21
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.7096774193548387,
|
| 181 |
+
"grad_norm": 0.15652534365653992,
|
| 182 |
+
"learning_rate": 1.990819274431662e-07,
|
| 183 |
+
"loss": 0.2952,
|
| 184 |
+
"mean_token_accuracy": 0.9242802858352661,
|
| 185 |
+
"step": 22
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.7419354838709677,
|
| 189 |
+
"grad_norm": 0.14285646378993988,
|
| 190 |
+
"learning_rate": 1.9875109281794824e-07,
|
| 191 |
+
"loss": 0.2787,
|
| 192 |
+
"mean_token_accuracy": 0.9304314851760864,
|
| 193 |
+
"step": 23
|
| 194 |
+
},
|
| 195 |
+
{
|
| 196 |
+
"epoch": 0.7741935483870968,
|
| 197 |
+
"grad_norm": 0.13358458876609802,
|
| 198 |
+
"learning_rate": 1.9836981604521074e-07,
|
| 199 |
+
"loss": 0.2705,
|
| 200 |
+
"mean_token_accuracy": 0.9316097497940063,
|
| 201 |
+
"step": 24
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 0.8064516129032258,
|
| 205 |
+
"grad_norm": 0.1293199211359024,
|
| 206 |
+
"learning_rate": 1.9793829188147403e-07,
|
| 207 |
+
"loss": 0.2975,
|
| 208 |
+
"mean_token_accuracy": 0.92156982421875,
|
| 209 |
+
"step": 25
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 0.8387096774193549,
|
| 213 |
+
"grad_norm": 0.13538742065429688,
|
| 214 |
+
"learning_rate": 1.9745674074967117e-07,
|
| 215 |
+
"loss": 0.2847,
|
| 216 |
+
"mean_token_accuracy": 0.9282350540161133,
|
| 217 |
+
"step": 26
|
| 218 |
+
},
|
| 219 |
+
{
|
| 220 |
+
"epoch": 0.8709677419354839,
|
| 221 |
+
"grad_norm": 0.146584153175354,
|
| 222 |
+
"learning_rate": 1.9692540862655585e-07,
|
| 223 |
+
"loss": 0.3159,
|
| 224 |
+
"mean_token_accuracy": 0.9188564419746399,
|
| 225 |
+
"step": 27
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 0.9032258064516129,
|
| 229 |
+
"grad_norm": 0.15393050014972687,
|
| 230 |
+
"learning_rate": 1.9634456691705702e-07,
|
| 231 |
+
"loss": 0.3065,
|
| 232 |
+
"mean_token_accuracy": 0.9200947284698486,
|
| 233 |
+
"step": 28
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.9354838709677419,
|
| 237 |
+
"grad_norm": 0.14639101922512054,
|
| 238 |
+
"learning_rate": 1.9571451231564522e-07,
|
| 239 |
+
"loss": 0.2877,
|
| 240 |
+
"mean_token_accuracy": 0.9255397319793701,
|
| 241 |
+
"step": 29
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.967741935483871,
|
| 245 |
+
"grad_norm": 0.14252065122127533,
|
| 246 |
+
"learning_rate": 1.9503556665478065e-07,
|
| 247 |
+
"loss": 0.3029,
|
| 248 |
+
"mean_token_accuracy": 0.9222911596298218,
|
| 249 |
+
"step": 30
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 1.0,
|
| 253 |
+
"grad_norm": 0.1796598732471466,
|
| 254 |
+
"learning_rate": 1.943080767405209e-07,
|
| 255 |
+
"loss": 0.3359,
|
| 256 |
+
"mean_token_accuracy": 0.9132773876190186,
|
| 257 |
+
"step": 31
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 1.032258064516129,
|
| 261 |
+
"grad_norm": 0.16374877095222473,
|
| 262 |
+
"learning_rate": 1.9353241417537212e-07,
|
| 263 |
+
"loss": 0.2893,
|
| 264 |
+
"mean_token_accuracy": 0.9236128926277161,
|
| 265 |
+
"step": 32
|
| 266 |
+
},
|
| 267 |
+
{
|
| 268 |
+
"epoch": 1.064516129032258,
|
| 269 |
+
"grad_norm": 0.1340431421995163,
|
| 270 |
+
"learning_rate": 1.9270897516847403e-07,
|
| 271 |
+
"loss": 0.2789,
|
| 272 |
+
"mean_token_accuracy": 0.9263336658477783,
|
| 273 |
+
"step": 33
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 1.096774193548387,
|
| 277 |
+
"grad_norm": 0.1781994253396988,
|
| 278 |
+
"learning_rate": 1.918381803332161e-07,
|
| 279 |
+
"loss": 0.3213,
|
| 280 |
+
"mean_token_accuracy": 0.9135346412658691,
|
| 281 |
+
"step": 34
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 1.129032258064516,
|
| 285 |
+
"grad_norm": 1900.6429443359375,
|
| 286 |
+
"learning_rate": 1.909204744723877e-07,
|
| 287 |
+
"loss": 0.2741,
|
| 288 |
+
"mean_token_accuracy": 0.9303540587425232,
|
| 289 |
+
"step": 35
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 1.1612903225806452,
|
| 293 |
+
"grad_norm": 0.16974130272865295,
|
| 294 |
+
"learning_rate": 1.8995632635097247e-07,
|
| 295 |
+
"loss": 0.3161,
|
| 296 |
+
"mean_token_accuracy": 0.9215410351753235,
|
| 297 |
+
"step": 36
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 1.1935483870967742,
|
| 301 |
+
"grad_norm": 0.14828237891197205,
|
| 302 |
+
"learning_rate": 1.889462284567028e-07,
|
| 303 |
+
"loss": 0.3141,
|
| 304 |
+
"mean_token_accuracy": 0.921214759349823,
|
| 305 |
+
"step": 37
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 1.2258064516129032,
|
| 309 |
+
"grad_norm": 992012608.0,
|
| 310 |
+
"learning_rate": 1.8789069674849658e-07,
|
| 311 |
+
"loss": 0.2945,
|
| 312 |
+
"mean_token_accuracy": 0.9253477454185486,
|
| 313 |
+
"step": 38
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 1.2580645161290323,
|
| 317 |
+
"grad_norm": 0.15507718920707703,
|
| 318 |
+
"learning_rate": 1.8679027039290496e-07,
|
| 319 |
+
"loss": 0.3063,
|
| 320 |
+
"mean_token_accuracy": 0.9237247705459595,
|
| 321 |
+
"step": 39
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 1.2903225806451613,
|
| 325 |
+
"grad_norm": 0.118597112596035,
|
| 326 |
+
"learning_rate": 1.856455114887056e-07,
|
| 327 |
+
"loss": 0.2595,
|
| 328 |
+
"mean_token_accuracy": 0.9301245808601379,
|
| 329 |
+
"step": 40
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 1.3225806451612903,
|
| 333 |
+
"grad_norm": 0.14978240430355072,
|
| 334 |
+
"learning_rate": 1.8445700477978204e-07,
|
| 335 |
+
"loss": 0.3051,
|
| 336 |
+
"mean_token_accuracy": 0.918502926826477,
|
| 337 |
+
"step": 41
|
| 338 |
+
},
|
| 339 |
+
{
|
| 340 |
+
"epoch": 1.3548387096774195,
|
| 341 |
+
"grad_norm": 0.14542274177074432,
|
| 342 |
+
"learning_rate": 1.8322535735643602e-07,
|
| 343 |
+
"loss": 0.2969,
|
| 344 |
+
"mean_token_accuracy": 0.91986483335495,
|
| 345 |
+
"step": 42
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 1.3870967741935485,
|
| 349 |
+
"grad_norm": 0.15233197808265686,
|
| 350 |
+
"learning_rate": 1.8195119834528532e-07,
|
| 351 |
+
"loss": 0.3005,
|
| 352 |
+
"mean_token_accuracy": 0.9208536148071289,
|
| 353 |
+
"step": 43
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 1.4193548387096775,
|
| 357 |
+
"grad_norm": 0.1700386255979538,
|
| 358 |
+
"learning_rate": 1.8063517858790515e-07,
|
| 359 |
+
"loss": 0.2974,
|
| 360 |
+
"mean_token_accuracy": 0.9260939359664917,
|
| 361 |
+
"step": 44
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 1.4516129032258065,
|
| 365 |
+
"grad_norm": 0.1596897840499878,
|
| 366 |
+
"learning_rate": 1.7927797030837767e-07,
|
| 367 |
+
"loss": 0.3092,
|
| 368 |
+
"mean_token_accuracy": 0.9239941239356995,
|
| 369 |
+
"step": 45
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 1.4838709677419355,
|
| 373 |
+
"grad_norm": 0.1388406604528427,
|
| 374 |
+
"learning_rate": 1.778802667699196e-07,
|
| 375 |
+
"loss": 0.2999,
|
| 376 |
+
"mean_token_accuracy": 0.9225243926048279,
|
| 377 |
+
"step": 46
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 1.5161290322580645,
|
| 381 |
+
"grad_norm": 591844999168.0,
|
| 382 |
+
"learning_rate": 1.764427819207624e-07,
|
| 383 |
+
"loss": 0.3303,
|
| 384 |
+
"mean_token_accuracy": 0.9121381044387817,
|
| 385 |
+
"step": 47
|
| 386 |
+
},
|
| 387 |
+
{
|
| 388 |
+
"epoch": 1.5483870967741935,
|
| 389 |
+
"grad_norm": 0.14387820661067963,
|
| 390 |
+
"learning_rate": 1.74966250029467e-07,
|
| 391 |
+
"loss": 0.3096,
|
| 392 |
+
"mean_token_accuracy": 0.9206516146659851,
|
| 393 |
+
"step": 48
|
| 394 |
+
},
|
| 395 |
+
{
|
| 396 |
+
"epoch": 1.5806451612903225,
|
| 397 |
+
"grad_norm": 0.1407996267080307,
|
| 398 |
+
"learning_rate": 1.7345142530985886e-07,
|
| 399 |
+
"loss": 0.2979,
|
| 400 |
+
"mean_token_accuracy": 0.9202678203582764,
|
| 401 |
+
"step": 49
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 1.6129032258064515,
|
| 405 |
+
"grad_norm": 0.13789159059524536,
|
| 406 |
+
"learning_rate": 1.718990815357747e-07,
|
| 407 |
+
"loss": 0.2723,
|
| 408 |
+
"mean_token_accuracy": 0.9288155436515808,
|
| 409 |
+
"step": 50
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 1.6451612903225805,
|
| 413 |
+
"grad_norm": 0.17413267493247986,
|
| 414 |
+
"learning_rate": 1.7031001164581827e-07,
|
| 415 |
+
"loss": 0.3462,
|
| 416 |
+
"mean_token_accuracy": 0.9114100933074951,
|
| 417 |
+
"step": 51
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 1.6774193548387095,
|
| 421 |
+
"grad_norm": 0.14843137562274933,
|
| 422 |
+
"learning_rate": 1.6868502733832642e-07,
|
| 423 |
+
"loss": 0.2886,
|
| 424 |
+
"mean_token_accuracy": 0.9235759973526001,
|
| 425 |
+
"step": 52
|
| 426 |
+
},
|
| 427 |
+
{
|
| 428 |
+
"epoch": 1.7096774193548387,
|
| 429 |
+
"grad_norm": 0.16399109363555908,
|
| 430 |
+
"learning_rate": 1.670249586567531e-07,
|
| 431 |
+
"loss": 0.3172,
|
| 432 |
+
"mean_token_accuracy": 0.9193130731582642,
|
| 433 |
+
"step": 53
|
| 434 |
+
},
|
| 435 |
+
{
|
| 436 |
+
"epoch": 1.7419354838709677,
|
| 437 |
+
"grad_norm": 0.14304129779338837,
|
| 438 |
+
"learning_rate": 1.6533065356568206e-07,
|
| 439 |
+
"loss": 0.2986,
|
| 440 |
+
"mean_token_accuracy": 0.9260614514350891,
|
| 441 |
+
"step": 54
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 1.7741935483870968,
|
| 445 |
+
"grad_norm": 0.15034537017345428,
|
| 446 |
+
"learning_rate": 1.636029775176862e-07,
|
| 447 |
+
"loss": 0.3142,
|
| 448 |
+
"mean_token_accuracy": 0.920920193195343,
|
| 449 |
+
"step": 55
|
| 450 |
+
},
|
| 451 |
+
{
|
| 452 |
+
"epoch": 1.8064516129032258,
|
| 453 |
+
"grad_norm": 0.13884751498699188,
|
| 454 |
+
"learning_rate": 1.618428130112533e-07,
|
| 455 |
+
"loss": 0.2859,
|
| 456 |
+
"mean_token_accuracy": 0.9274208545684814,
|
| 457 |
+
"step": 56
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 1.838709677419355,
|
| 461 |
+
"grad_norm": 0.14844031631946564,
|
| 462 |
+
"learning_rate": 1.6005105914000505e-07,
|
| 463 |
+
"loss": 0.2981,
|
| 464 |
+
"mean_token_accuracy": 0.9255636930465698,
|
| 465 |
+
"step": 57
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 1.870967741935484,
|
| 469 |
+
"grad_norm": 0.16273614764213562,
|
| 470 |
+
"learning_rate": 1.5822863113343934e-07,
|
| 471 |
+
"loss": 0.3317,
|
| 472 |
+
"mean_token_accuracy": 0.9183942079544067,
|
| 473 |
+
"step": 58
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 1.903225806451613,
|
| 477 |
+
"grad_norm": 2596124672.0,
|
| 478 |
+
"learning_rate": 1.5637645988943006e-07,
|
| 479 |
+
"loss": 0.3263,
|
| 480 |
+
"mean_token_accuracy": 0.9169372320175171,
|
| 481 |
+
"step": 59
|
| 482 |
+
},
|
| 483 |
+
{
|
| 484 |
+
"epoch": 1.935483870967742,
|
| 485 |
+
"grad_norm": 0.15882141888141632,
|
| 486 |
+
"learning_rate": 1.5449549149872375e-07,
|
| 487 |
+
"loss": 0.2856,
|
| 488 |
+
"mean_token_accuracy": 0.9287616610527039,
|
| 489 |
+
"step": 60
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 1.967741935483871,
|
| 493 |
+
"grad_norm": 0.13461528718471527,
|
| 494 |
+
"learning_rate": 1.5258668676167547e-07,
|
| 495 |
+
"loss": 0.2996,
|
| 496 |
+
"mean_token_accuracy": 0.9221404790878296,
|
| 497 |
+
"step": 61
|
| 498 |
+
},
|
| 499 |
+
{
|
| 500 |
+
"epoch": 2.0,
|
| 501 |
+
"grad_norm": 0.17273001372814178,
|
| 502 |
+
"learning_rate": 1.5065102069747116e-07,
|
| 503 |
+
"loss": 0.3115,
|
| 504 |
+
"mean_token_accuracy": 0.9187781810760498,
|
| 505 |
+
"step": 62
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"epoch": 2.032258064516129,
|
| 509 |
+
"grad_norm": 0.13375335931777954,
|
| 510 |
+
"learning_rate": 1.4868948204608697e-07,
|
| 511 |
+
"loss": 0.2721,
|
| 512 |
+
"mean_token_accuracy": 0.9304314851760864,
|
| 513 |
+
"step": 63
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 2.064516129032258,
|
| 517 |
+
"grad_norm": 0.13471315801143646,
|
| 518 |
+
"learning_rate": 1.4670307276324006e-07,
|
| 519 |
+
"loss": 0.299,
|
| 520 |
+
"mean_token_accuracy": 0.9216856360435486,
|
| 521 |
+
"step": 64
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 2.096774193548387,
|
| 525 |
+
"grad_norm": 0.12814444303512573,
|
| 526 |
+
"learning_rate": 1.4469280750858852e-07,
|
| 527 |
+
"loss": 0.296,
|
| 528 |
+
"mean_token_accuracy": 0.9213560819625854,
|
| 529 |
+
"step": 65
|
| 530 |
+
},
|
| 531 |
+
{
|
| 532 |
+
"epoch": 2.129032258064516,
|
| 533 |
+
"grad_norm": 0.14648547768592834,
|
| 534 |
+
"learning_rate": 1.4265971312744249e-07,
|
| 535 |
+
"loss": 0.2833,
|
| 536 |
+
"mean_token_accuracy": 0.9275593757629395,
|
| 537 |
+
"step": 66
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 2.161290322580645,
|
| 541 |
+
"grad_norm": 0.1619262844324112,
|
| 542 |
+
"learning_rate": 1.4060482812625054e-07,
|
| 543 |
+
"loss": 0.3123,
|
| 544 |
+
"mean_token_accuracy": 0.9240583181381226,
|
| 545 |
+
"step": 67
|
| 546 |
+
},
|
| 547 |
+
{
|
| 548 |
+
"epoch": 2.193548387096774,
|
| 549 |
+
"grad_norm": 0.15138781070709229,
|
| 550 |
+
"learning_rate": 1.3852920214212964e-07,
|
| 551 |
+
"loss": 0.3084,
|
| 552 |
+
"mean_token_accuracy": 0.9199481010437012,
|
| 553 |
+
"step": 68
|
| 554 |
+
},
|
| 555 |
+
{
|
| 556 |
+
"epoch": 2.225806451612903,
|
| 557 |
+
"grad_norm": 0.18671105802059174,
|
| 558 |
+
"learning_rate": 1.3643389540670962e-07,
|
| 559 |
+
"loss": 0.3394,
|
| 560 |
+
"mean_token_accuracy": 0.9127894639968872,
|
| 561 |
+
"step": 69
|
| 562 |
+
},
|
| 563 |
+
{
|
| 564 |
+
"epoch": 2.258064516129032,
|
| 565 |
+
"grad_norm": 0.12655843794345856,
|
| 566 |
+
"learning_rate": 1.3431997820456592e-07,
|
| 567 |
+
"loss": 0.3004,
|
| 568 |
+
"mean_token_accuracy": 0.9205244779586792,
|
| 569 |
+
"step": 70
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 2.2903225806451615,
|
| 573 |
+
"grad_norm": 0.1643962413072586,
|
| 574 |
+
"learning_rate": 1.3218853032651718e-07,
|
| 575 |
+
"loss": 0.2982,
|
| 576 |
+
"mean_token_accuracy": 0.9248165488243103,
|
| 577 |
+
"step": 71
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 2.3225806451612905,
|
| 581 |
+
"grad_norm": 0.1438770592212677,
|
| 582 |
+
"learning_rate": 1.300406405180671e-07,
|
| 583 |
+
"loss": 0.2893,
|
| 584 |
+
"mean_token_accuracy": 0.9276708960533142,
|
| 585 |
+
"step": 72
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 2.3548387096774195,
|
| 589 |
+
"grad_norm": 13705685.0,
|
| 590 |
+
"learning_rate": 1.278774059232723e-07,
|
| 591 |
+
"loss": 0.2693,
|
| 592 |
+
"mean_token_accuracy": 0.9281597137451172,
|
| 593 |
+
"step": 73
|
| 594 |
+
},
|
| 595 |
+
{
|
| 596 |
+
"epoch": 2.3870967741935485,
|
| 597 |
+
"grad_norm": 0.1379764974117279,
|
| 598 |
+
"learning_rate": 1.2569993152432026e-07,
|
| 599 |
+
"loss": 0.3048,
|
| 600 |
+
"mean_token_accuracy": 0.9185689091682434,
|
| 601 |
+
"step": 74
|
| 602 |
+
},
|
| 603 |
+
{
|
| 604 |
+
"epoch": 2.4193548387096775,
|
| 605 |
+
"grad_norm": 0.1545405089855194,
|
| 606 |
+
"learning_rate": 1.2350932957710321e-07,
|
| 607 |
+
"loss": 0.2992,
|
| 608 |
+
"mean_token_accuracy": 0.921677827835083,
|
| 609 |
+
"step": 75
|
| 610 |
+
},
|
| 611 |
+
{
|
| 612 |
+
"epoch": 2.4516129032258065,
|
| 613 |
+
"grad_norm": 0.16681832075119019,
|
| 614 |
+
"learning_rate": 1.213067190430769e-07,
|
| 615 |
+
"loss": 0.3072,
|
| 616 |
+
"mean_token_accuracy": 0.9232601523399353,
|
| 617 |
+
"step": 76
|
| 618 |
+
},
|
| 619 |
+
{
|
| 620 |
+
"epoch": 2.4838709677419355,
|
| 621 |
+
"grad_norm": 0.16876104474067688,
|
| 622 |
+
"learning_rate": 1.1909322501769406e-07,
|
| 623 |
+
"loss": 0.3232,
|
| 624 |
+
"mean_token_accuracy": 0.9157925844192505,
|
| 625 |
+
"step": 77
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 2.5161290322580645,
|
| 629 |
+
"grad_norm": 0.14914129674434662,
|
| 630 |
+
"learning_rate": 1.1686997815570472e-07,
|
| 631 |
+
"loss": 0.3022,
|
| 632 |
+
"mean_token_accuracy": 0.9226891994476318,
|
| 633 |
+
"step": 78
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 2.5483870967741935,
|
| 637 |
+
"grad_norm": 0.12800626456737518,
|
| 638 |
+
"learning_rate": 1.1463811409361665e-07,
|
| 639 |
+
"loss": 0.3029,
|
| 640 |
+
"mean_token_accuracy": 0.9234021902084351,
|
| 641 |
+
"step": 79
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"epoch": 2.5806451612903225,
|
| 645 |
+
"grad_norm": 0.16127674281597137,
|
| 646 |
+
"learning_rate": 1.1239877286961121e-07,
|
| 647 |
+
"loss": 0.313,
|
| 648 |
+
"mean_token_accuracy": 0.9184219837188721,
|
| 649 |
+
"step": 80
|
| 650 |
+
},
|
| 651 |
+
{
|
| 652 |
+
"epoch": 2.6129032258064515,
|
| 653 |
+
"grad_norm": 397858209792.0,
|
| 654 |
+
"learning_rate": 1.101530983412108e-07,
|
| 655 |
+
"loss": 0.3366,
|
| 656 |
+
"mean_token_accuracy": 0.9118554592132568,
|
| 657 |
+
"step": 81
|
| 658 |
+
},
|
| 659 |
+
{
|
| 660 |
+
"epoch": 2.6451612903225805,
|
| 661 |
+
"grad_norm": 107614896128.0,
|
| 662 |
+
"learning_rate": 1.0790223760099548e-07,
|
| 663 |
+
"loss": 0.2926,
|
| 664 |
+
"mean_token_accuracy": 0.92240309715271,
|
| 665 |
+
"step": 82
|
| 666 |
+
},
|
| 667 |
+
{
|
| 668 |
+
"epoch": 2.6774193548387095,
|
| 669 |
+
"grad_norm": 0.12381122261285782,
|
| 670 |
+
"learning_rate": 1.0564734039066698e-07,
|
| 671 |
+
"loss": 0.2736,
|
| 672 |
+
"mean_token_accuracy": 0.9273778796195984,
|
| 673 |
+
"step": 83
|
| 674 |
+
},
|
| 675 |
+
{
|
| 676 |
+
"epoch": 2.709677419354839,
|
| 677 |
+
"grad_norm": 0.17293043434619904,
|
| 678 |
+
"learning_rate": 1.0338955851375961e-07,
|
| 679 |
+
"loss": 0.3302,
|
| 680 |
+
"mean_token_accuracy": 0.9148573875427246,
|
| 681 |
+
"step": 84
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 2.741935483870968,
|
| 685 |
+
"grad_norm": 318356848640.0,
|
| 686 |
+
"learning_rate": 1.0113004524729798e-07,
|
| 687 |
+
"loss": 0.2887,
|
| 688 |
+
"mean_token_accuracy": 0.9243521094322205,
|
| 689 |
+
"step": 85
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 2.774193548387097,
|
| 693 |
+
"grad_norm": 0.13202118873596191,
|
| 694 |
+
"learning_rate": 9.886995475270203e-08,
|
| 695 |
+
"loss": 0.2765,
|
| 696 |
+
"mean_token_accuracy": 0.9281702637672424,
|
| 697 |
+
"step": 86
|
| 698 |
+
},
|
| 699 |
+
{
|
| 700 |
+
"epoch": 2.806451612903226,
|
| 701 |
+
"grad_norm": 0.13786709308624268,
|
| 702 |
+
"learning_rate": 9.661044148624036e-08,
|
| 703 |
+
"loss": 0.2712,
|
| 704 |
+
"mean_token_accuracy": 0.9306640625,
|
| 705 |
+
"step": 87
|
| 706 |
+
},
|
| 707 |
+
{
|
| 708 |
+
"epoch": 2.838709677419355,
|
| 709 |
+
"grad_norm": 0.1563229262828827,
|
| 710 |
+
"learning_rate": 9.435265960933302e-08,
|
| 711 |
+
"loss": 0.2965,
|
| 712 |
+
"mean_token_accuracy": 0.9226992130279541,
|
| 713 |
+
"step": 88
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 2.870967741935484,
|
| 717 |
+
"grad_norm": 0.16793645918369293,
|
| 718 |
+
"learning_rate": 9.209776239900452e-08,
|
| 719 |
+
"loss": 0.3205,
|
| 720 |
+
"mean_token_accuracy": 0.9166249632835388,
|
| 721 |
+
"step": 89
|
| 722 |
+
},
|
| 723 |
+
{
|
| 724 |
+
"epoch": 2.903225806451613,
|
| 725 |
+
"grad_norm": 0.19805242121219635,
|
| 726 |
+
"learning_rate": 8.98469016587892e-08,
|
| 727 |
+
"loss": 0.3144,
|
| 728 |
+
"mean_token_accuracy": 0.9230661392211914,
|
| 729 |
+
"step": 90
|
| 730 |
+
},
|
| 731 |
+
{
|
| 732 |
+
"epoch": 2.935483870967742,
|
| 733 |
+
"grad_norm": 0.17880718410015106,
|
| 734 |
+
"learning_rate": 8.76012271303888e-08,
|
| 735 |
+
"loss": 0.3426,
|
| 736 |
+
"mean_token_accuracy": 0.9152123332023621,
|
| 737 |
+
"step": 91
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 2.967741935483871,
|
| 741 |
+
"grad_norm": 0.13014861941337585,
|
| 742 |
+
"learning_rate": 8.536188590638333e-08,
|
| 743 |
+
"loss": 0.2849,
|
| 744 |
+
"mean_token_accuracy": 0.9278050065040588,
|
| 745 |
+
"step": 92
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 3.0,
|
| 749 |
+
"grad_norm": 134254736.0,
|
| 750 |
+
"learning_rate": 8.313002184429528e-08,
|
| 751 |
+
"loss": 0.3084,
|
| 752 |
+
"mean_token_accuracy": 0.922960638999939,
|
| 753 |
+
"step": 93
|
| 754 |
+
},
|
| 755 |
+
{
|
| 756 |
+
"epoch": 3.032258064516129,
|
| 757 |
+
"grad_norm": 2311489060864.0,
|
| 758 |
+
"learning_rate": 8.090677498230596e-08,
|
| 759 |
+
"loss": 0.3011,
|
| 760 |
+
"mean_token_accuracy": 0.919518232345581,
|
| 761 |
+
"step": 94
|
| 762 |
+
},
|
| 763 |
+
{
|
| 764 |
+
"epoch": 3.064516129032258,
|
| 765 |
+
"grad_norm": 0.1483820080757141,
|
| 766 |
+
"learning_rate": 7.869328095692311e-08,
|
| 767 |
+
"loss": 0.3036,
|
| 768 |
+
"mean_token_accuracy": 0.9224164485931396,
|
| 769 |
+
"step": 95
|
| 770 |
+
},
|
| 771 |
+
{
|
| 772 |
+
"epoch": 3.096774193548387,
|
| 773 |
+
"grad_norm": 0.1287539154291153,
|
| 774 |
+
"learning_rate": 7.64906704228968e-08,
|
| 775 |
+
"loss": 0.3109,
|
| 776 |
+
"mean_token_accuracy": 0.9177184104919434,
|
| 777 |
+
"step": 96
|
| 778 |
+
},
|
| 779 |
+
{
|
| 780 |
+
"epoch": 3.129032258064516,
|
| 781 |
+
"grad_norm": 0.14654560387134552,
|
| 782 |
+
"learning_rate": 7.43000684756797e-08,
|
| 783 |
+
"loss": 0.2991,
|
| 784 |
+
"mean_token_accuracy": 0.9215620756149292,
|
| 785 |
+
"step": 97
|
| 786 |
+
},
|
| 787 |
+
{
|
| 788 |
+
"epoch": 3.161290322580645,
|
| 789 |
+
"grad_norm": 0.15880532562732697,
|
| 790 |
+
"learning_rate": 7.21225940767277e-08,
|
| 791 |
+
"loss": 0.2717,
|
| 792 |
+
"mean_token_accuracy": 0.9311442971229553,
|
| 793 |
+
"step": 98
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 3.193548387096774,
|
| 797 |
+
"grad_norm": 0.15332385897636414,
|
| 798 |
+
"learning_rate": 6.995935948193294e-08,
|
| 799 |
+
"loss": 0.2906,
|
| 800 |
+
"mean_token_accuracy": 0.9262657761573792,
|
| 801 |
+
"step": 99
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 3.225806451612903,
|
| 805 |
+
"grad_norm": 0.16386571526527405,
|
| 806 |
+
"learning_rate": 6.781146967348282e-08,
|
| 807 |
+
"loss": 0.3087,
|
| 808 |
+
"mean_token_accuracy": 0.9201537370681763,
|
| 809 |
+
"step": 100
|
| 810 |
+
},
|
| 811 |
+
{
|
| 812 |
+
"epoch": 3.258064516129032,
|
| 813 |
+
"grad_norm": 0.1456972062587738,
|
| 814 |
+
"learning_rate": 6.568002179543408e-08,
|
| 815 |
+
"loss": 0.2845,
|
| 816 |
+
"mean_token_accuracy": 0.9251313209533691,
|
| 817 |
+
"step": 101
|
| 818 |
+
},
|
| 819 |
+
{
|
| 820 |
+
"epoch": 3.2903225806451615,
|
| 821 |
+
"grad_norm": 0.1516929566860199,
|
| 822 |
+
"learning_rate": 6.356610459329037e-08,
|
| 823 |
+
"loss": 0.312,
|
| 824 |
+
"mean_token_accuracy": 0.9195959568023682,
|
| 825 |
+
"step": 102
|
| 826 |
+
},
|
| 827 |
+
{
|
| 828 |
+
"epoch": 3.3225806451612905,
|
| 829 |
+
"grad_norm": 5110276608.0,
|
| 830 |
+
"learning_rate": 6.147079785787038e-08,
|
| 831 |
+
"loss": 0.3149,
|
| 832 |
+
"mean_token_accuracy": 0.9239295125007629,
|
| 833 |
+
"step": 103
|
| 834 |
+
},
|
| 835 |
+
{
|
| 836 |
+
"epoch": 3.3548387096774195,
|
| 837 |
+
"grad_norm": 0.14113079011440277,
|
| 838 |
+
"learning_rate": 5.939517187374949e-08,
|
| 839 |
+
"loss": 0.31,
|
| 840 |
+
"mean_token_accuracy": 0.9190698862075806,
|
| 841 |
+
"step": 104
|
| 842 |
+
},
|
| 843 |
+
{
|
| 844 |
+
"epoch": 3.3870967741935485,
|
| 845 |
+
"grad_norm": 0.15635834634304047,
|
| 846 |
+
"learning_rate": 5.7340286872557505e-08,
|
| 847 |
+
"loss": 0.3075,
|
| 848 |
+
"mean_token_accuracy": 0.9230503439903259,
|
| 849 |
+
"step": 105
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 3.4193548387096775,
|
| 853 |
+
"grad_norm": 0.15426763892173767,
|
| 854 |
+
"learning_rate": 5.530719249141147e-08,
|
| 855 |
+
"loss": 0.3232,
|
| 856 |
+
"mean_token_accuracy": 0.919147253036499,
|
| 857 |
+
"step": 106
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 3.4516129032258065,
|
| 861 |
+
"grad_norm": 0.1907190978527069,
|
| 862 |
+
"learning_rate": 5.3296927236759934e-08,
|
| 863 |
+
"loss": 0.3244,
|
| 864 |
+
"mean_token_accuracy": 0.914537787437439,
|
| 865 |
+
"step": 107
|
| 866 |
+
},
|
| 867 |
+
{
|
| 868 |
+
"epoch": 3.4838709677419355,
|
| 869 |
+
"grad_norm": 0.14416781067848206,
|
| 870 |
+
"learning_rate": 5.131051795391301e-08,
|
| 871 |
+
"loss": 0.3002,
|
| 872 |
+
"mean_token_accuracy": 0.924168050289154,
|
| 873 |
+
"step": 108
|
| 874 |
+
},
|
| 875 |
+
{
|
| 876 |
+
"epoch": 3.5161290322580645,
|
| 877 |
+
"grad_norm": 0.16144634783267975,
|
| 878 |
+
"learning_rate": 4.934897930252886e-08,
|
| 879 |
+
"loss": 0.3025,
|
| 880 |
+
"mean_token_accuracy": 0.9218240976333618,
|
| 881 |
+
"step": 109
|
| 882 |
+
},
|
| 883 |
+
{
|
| 884 |
+
"epoch": 3.5483870967741935,
|
| 885 |
+
"grad_norm": 0.15450312197208405,
|
| 886 |
+
"learning_rate": 4.741331323832455e-08,
|
| 887 |
+
"loss": 0.2909,
|
| 888 |
+
"mean_token_accuracy": 0.924649178981781,
|
| 889 |
+
"step": 110
|
| 890 |
+
},
|
| 891 |
+
{
|
| 892 |
+
"epoch": 3.5806451612903225,
|
| 893 |
+
"grad_norm": 0.12900131940841675,
|
| 894 |
+
"learning_rate": 4.5504508501276254e-08,
|
| 895 |
+
"loss": 0.2798,
|
| 896 |
+
"mean_token_accuracy": 0.9257345795631409,
|
| 897 |
+
"step": 111
|
| 898 |
+
},
|
| 899 |
+
{
|
| 900 |
+
"epoch": 3.6129032258064515,
|
| 901 |
+
"grad_norm": 0.14584125578403473,
|
| 902 |
+
"learning_rate": 4.3623540110569934e-08,
|
| 903 |
+
"loss": 0.308,
|
| 904 |
+
"mean_token_accuracy": 0.9198503494262695,
|
| 905 |
+
"step": 112
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 3.6451612903225805,
|
| 909 |
+
"grad_norm": 0.13434411585330963,
|
| 910 |
+
"learning_rate": 4.1771368866560665e-08,
|
| 911 |
+
"loss": 0.2937,
|
| 912 |
+
"mean_token_accuracy": 0.9217662215232849,
|
| 913 |
+
"step": 113
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 3.6774193548387095,
|
| 917 |
+
"grad_norm": 0.15536189079284668,
|
| 918 |
+
"learning_rate": 3.9948940859994963e-08,
|
| 919 |
+
"loss": 0.3124,
|
| 920 |
+
"mean_token_accuracy": 0.9192003011703491,
|
| 921 |
+
"step": 114
|
| 922 |
+
},
|
| 923 |
+
{
|
| 924 |
+
"epoch": 3.709677419354839,
|
| 925 |
+
"grad_norm": 0.16448096930980682,
|
| 926 |
+
"learning_rate": 3.8157186988746716e-08,
|
| 927 |
+
"loss": 0.2935,
|
| 928 |
+
"mean_token_accuracy": 0.927962064743042,
|
| 929 |
+
"step": 115
|
| 930 |
+
},
|
| 931 |
+
{
|
| 932 |
+
"epoch": 3.741935483870968,
|
| 933 |
+
"grad_norm": 0.18926770985126495,
|
| 934 |
+
"learning_rate": 3.63970224823138e-08,
|
| 935 |
+
"loss": 0.3329,
|
| 936 |
+
"mean_token_accuracy": 0.9149773716926575,
|
| 937 |
+
"step": 116
|
| 938 |
+
},
|
| 939 |
+
{
|
| 940 |
+
"epoch": 3.774193548387097,
|
| 941 |
+
"grad_norm": 0.14664557576179504,
|
| 942 |
+
"learning_rate": 3.4669346434317946e-08,
|
| 943 |
+
"loss": 0.2902,
|
| 944 |
+
"mean_token_accuracy": 0.9280288219451904,
|
| 945 |
+
"step": 117
|
| 946 |
+
},
|
| 947 |
+
{
|
| 948 |
+
"epoch": 3.806451612903226,
|
| 949 |
+
"grad_norm": 0.17194537818431854,
|
| 950 |
+
"learning_rate": 3.297504134324693e-08,
|
| 951 |
+
"loss": 0.306,
|
| 952 |
+
"mean_token_accuracy": 0.922054648399353,
|
| 953 |
+
"step": 118
|
| 954 |
+
},
|
| 955 |
+
{
|
| 956 |
+
"epoch": 3.838709677419355,
|
| 957 |
+
"grad_norm": 0.13546685874462128,
|
| 958 |
+
"learning_rate": 3.131497266167357e-08,
|
| 959 |
+
"loss": 0.2662,
|
| 960 |
+
"mean_token_accuracy": 0.933337390422821,
|
| 961 |
+
"step": 119
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 3.870967741935484,
|
| 965 |
+
"grad_norm": 0.15774165093898773,
|
| 966 |
+
"learning_rate": 2.9689988354181737e-08,
|
| 967 |
+
"loss": 0.3042,
|
| 968 |
+
"mean_token_accuracy": 0.9215238094329834,
|
| 969 |
+
"step": 120
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 3.903225806451613,
|
| 973 |
+
"grad_norm": 11272906080256.0,
|
| 974 |
+
"learning_rate": 2.81009184642253e-08,
|
| 975 |
+
"loss": 0.2931,
|
| 976 |
+
"mean_token_accuracy": 0.9249425530433655,
|
| 977 |
+
"step": 121
|
| 978 |
+
},
|
| 979 |
+
{
|
| 980 |
+
"epoch": 3.935483870967742,
|
| 981 |
+
"grad_norm": 0.13712410628795624,
|
| 982 |
+
"learning_rate": 2.6548574690141122e-08,
|
| 983 |
+
"loss": 0.291,
|
| 984 |
+
"mean_token_accuracy": 0.9214818477630615,
|
| 985 |
+
"step": 122
|
| 986 |
+
},
|
| 987 |
+
{
|
| 988 |
+
"epoch": 3.967741935483871,
|
| 989 |
+
"grad_norm": 0.1492443084716797,
|
| 990 |
+
"learning_rate": 2.5033749970533015e-08,
|
| 991 |
+
"loss": 0.3333,
|
| 992 |
+
"mean_token_accuracy": 0.91297447681427,
|
| 993 |
+
"step": 123
|
| 994 |
+
},
|
| 995 |
+
{
|
| 996 |
+
"epoch": 4.0,
|
| 997 |
+
"grad_norm": 0.14738278090953827,
|
| 998 |
+
"learning_rate": 2.3557218079237607e-08,
|
| 999 |
+
"loss": 0.3008,
|
| 1000 |
+
"mean_token_accuracy": 0.9269067049026489,
|
| 1001 |
+
"step": 124
|
| 1002 |
+
},
|
| 1003 |
+
{
|
| 1004 |
+
"epoch": 4.032258064516129,
|
| 1005 |
+
"grad_norm": 0.14828868210315704,
|
| 1006 |
+
"learning_rate": 2.2119733230080406e-08,
|
| 1007 |
+
"loss": 0.3025,
|
| 1008 |
+
"mean_token_accuracy": 0.918917179107666,
|
| 1009 |
+
"step": 125
|
| 1010 |
+
},
|
| 1011 |
+
{
|
| 1012 |
+
"epoch": 4.064516129032258,
|
| 1013 |
+
"grad_norm": 0.129131019115448,
|
| 1014 |
+
"learning_rate": 2.0722029691622334e-08,
|
| 1015 |
+
"loss": 0.2981,
|
| 1016 |
+
"mean_token_accuracy": 0.9248138666152954,
|
| 1017 |
+
"step": 126
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 4.096774193548387,
|
| 1021 |
+
"grad_norm": 0.16759882867336273,
|
| 1022 |
+
"learning_rate": 1.9364821412094857e-08,
|
| 1023 |
+
"loss": 0.3119,
|
| 1024 |
+
"mean_token_accuracy": 0.9213570952415466,
|
| 1025 |
+
"step": 127
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 4.129032258064516,
|
| 1029 |
+
"grad_norm": 0.16001583635807037,
|
| 1030 |
+
"learning_rate": 1.8048801654714683e-08,
|
| 1031 |
+
"loss": 0.3155,
|
| 1032 |
+
"mean_token_accuracy": 0.920055627822876,
|
| 1033 |
+
"step": 128
|
| 1034 |
+
},
|
| 1035 |
+
{
|
| 1036 |
+
"epoch": 4.161290322580645,
|
| 1037 |
+
"grad_norm": 0.14847075939178467,
|
| 1038 |
+
"learning_rate": 1.677464264356395e-08,
|
| 1039 |
+
"loss": 0.3099,
|
| 1040 |
+
"mean_token_accuracy": 0.9192079305648804,
|
| 1041 |
+
"step": 129
|
| 1042 |
+
},
|
| 1043 |
+
{
|
| 1044 |
+
"epoch": 4.193548387096774,
|
| 1045 |
+
"grad_norm": 0.16150416433811188,
|
| 1046 |
+
"learning_rate": 1.554299522021796e-08,
|
| 1047 |
+
"loss": 0.3107,
|
| 1048 |
+
"mean_token_accuracy": 0.9221417903900146,
|
| 1049 |
+
"step": 130
|
| 1050 |
+
},
|
| 1051 |
+
{
|
| 1052 |
+
"epoch": 4.225806451612903,
|
| 1053 |
+
"grad_norm": 0.159551203250885,
|
| 1054 |
+
"learning_rate": 1.4354488511294416e-08,
|
| 1055 |
+
"loss": 0.3024,
|
| 1056 |
+
"mean_token_accuracy": 0.9230970144271851,
|
| 1057 |
+
"step": 131
|
| 1058 |
+
},
|
| 1059 |
+
{
|
| 1060 |
+
"epoch": 4.258064516129032,
|
| 1061 |
+
"grad_norm": 0.13861845433712006,
|
| 1062 |
+
"learning_rate": 1.3209729607095021e-08,
|
| 1063 |
+
"loss": 0.2794,
|
| 1064 |
+
"mean_token_accuracy": 0.9251081347465515,
|
| 1065 |
+
"step": 132
|
| 1066 |
+
},
|
| 1067 |
+
{
|
| 1068 |
+
"epoch": 4.290322580645161,
|
| 1069 |
+
"grad_norm": 0.1300177425146103,
|
| 1070 |
+
"learning_rate": 1.2109303251503433e-08,
|
| 1071 |
+
"loss": 0.2767,
|
| 1072 |
+
"mean_token_accuracy": 0.9269058108329773,
|
| 1073 |
+
"step": 133
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 4.32258064516129,
|
| 1077 |
+
"grad_norm": 0.15271711349487305,
|
| 1078 |
+
"learning_rate": 1.1053771543297197e-08,
|
| 1079 |
+
"loss": 0.2833,
|
| 1080 |
+
"mean_token_accuracy": 0.9278181791305542,
|
| 1081 |
+
"step": 134
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 4.354838709677419,
|
| 1085 |
+
"grad_norm": 0.15096405148506165,
|
| 1086 |
+
"learning_rate": 1.0043673649027518e-08,
|
| 1087 |
+
"loss": 0.3032,
|
| 1088 |
+
"mean_token_accuracy": 0.9209933876991272,
|
| 1089 |
+
"step": 135
|
| 1090 |
+
},
|
| 1091 |
+
{
|
| 1092 |
+
"epoch": 4.387096774193548,
|
| 1093 |
+
"grad_norm": 0.1546061486005783,
|
| 1094 |
+
"learning_rate": 9.07952552761232e-09,
|
| 1095 |
+
"loss": 0.2895,
|
| 1096 |
+
"mean_token_accuracy": 0.923534095287323,
|
| 1097 |
+
"step": 136
|
| 1098 |
+
},
|
| 1099 |
+
{
|
| 1100 |
+
"epoch": 4.419354838709677,
|
| 1101 |
+
"grad_norm": 0.15047194063663483,
|
| 1102 |
+
"learning_rate": 8.161819666783887e-09,
|
| 1103 |
+
"loss": 0.302,
|
| 1104 |
+
"mean_token_accuracy": 0.9207534790039062,
|
| 1105 |
+
"step": 137
|
| 1106 |
+
},
|
| 1107 |
+
{
|
| 1108 |
+
"epoch": 4.451612903225806,
|
| 1109 |
+
"grad_norm": 0.13835148513317108,
|
| 1110 |
+
"learning_rate": 7.29102483152596e-09,
|
| 1111 |
+
"loss": 0.2829,
|
| 1112 |
+
"mean_token_accuracy": 0.9299312829971313,
|
| 1113 |
+
"step": 138
|
| 1114 |
+
},
|
| 1115 |
+
{
|
| 1116 |
+
"epoch": 4.483870967741936,
|
| 1117 |
+
"grad_norm": 0.14567099511623383,
|
| 1118 |
+
"learning_rate": 6.467585824627886e-09,
|
| 1119 |
+
"loss": 0.3009,
|
| 1120 |
+
"mean_token_accuracy": 0.92234206199646,
|
| 1121 |
+
"step": 139
|
| 1122 |
+
},
|
| 1123 |
+
{
|
| 1124 |
+
"epoch": 4.516129032258064,
|
| 1125 |
+
"grad_norm": 0.14242888987064362,
|
| 1126 |
+
"learning_rate": 5.691923259479092e-09,
|
| 1127 |
+
"loss": 0.2946,
|
| 1128 |
+
"mean_token_accuracy": 0.9222550988197327,
|
| 1129 |
+
"step": 140
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 4.548387096774194,
|
| 1133 |
+
"grad_norm": 0.14662319421768188,
|
| 1134 |
+
"learning_rate": 4.964433345219354e-09,
|
| 1135 |
+
"loss": 0.3089,
|
| 1136 |
+
"mean_token_accuracy": 0.9191495180130005,
|
| 1137 |
+
"step": 141
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 4.580645161290323,
|
| 1141 |
+
"grad_norm": 0.16040538251399994,
|
| 1142 |
+
"learning_rate": 4.285487684354771e-09,
|
| 1143 |
+
"loss": 0.3086,
|
| 1144 |
+
"mean_token_accuracy": 0.9262273907661438,
|
| 1145 |
+
"step": 142
|
| 1146 |
+
},
|
| 1147 |
+
{
|
| 1148 |
+
"epoch": 4.612903225806452,
|
| 1149 |
+
"grad_norm": 0.17762604355812073,
|
| 1150 |
+
"learning_rate": 3.6554330829429714e-09,
|
| 1151 |
+
"loss": 0.3128,
|
| 1152 |
+
"mean_token_accuracy": 0.9169263243675232,
|
| 1153 |
+
"step": 143
|
| 1154 |
+
},
|
| 1155 |
+
{
|
| 1156 |
+
"epoch": 4.645161290322581,
|
| 1157 |
+
"grad_norm": 0.15534666180610657,
|
| 1158 |
+
"learning_rate": 3.074591373444135e-09,
|
| 1159 |
+
"loss": 0.3014,
|
| 1160 |
+
"mean_token_accuracy": 0.9234717488288879,
|
| 1161 |
+
"step": 144
|
| 1162 |
+
},
|
| 1163 |
+
{
|
| 1164 |
+
"epoch": 4.67741935483871,
|
| 1165 |
+
"grad_norm": 0.13566385209560394,
|
| 1166 |
+
"learning_rate": 2.5432592503287997e-09,
|
| 1167 |
+
"loss": 0.2719,
|
| 1168 |
+
"mean_token_accuracy": 0.9335603713989258,
|
| 1169 |
+
"step": 145
|
| 1170 |
+
},
|
| 1171 |
+
{
|
| 1172 |
+
"epoch": 4.709677419354839,
|
| 1173 |
+
"grad_norm": 0.1675826907157898,
|
| 1174 |
+
"learning_rate": 2.061708118525951e-09,
|
| 1175 |
+
"loss": 0.3175,
|
| 1176 |
+
"mean_token_accuracy": 0.925308108329773,
|
| 1177 |
+
"step": 146
|
| 1178 |
+
},
|
| 1179 |
+
{
|
| 1180 |
+
"epoch": 4.741935483870968,
|
| 1181 |
+
"grad_norm": 0.1770351529121399,
|
| 1182 |
+
"learning_rate": 1.6301839547892327e-09,
|
| 1183 |
+
"loss": 0.3083,
|
| 1184 |
+
"mean_token_accuracy": 0.9258779883384705,
|
| 1185 |
+
"step": 147
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 4.774193548387097,
|
| 1189 |
+
"grad_norm": 0.1430787742137909,
|
| 1190 |
+
"learning_rate": 1.2489071820517394e-09,
|
| 1191 |
+
"loss": 0.2772,
|
| 1192 |
+
"mean_token_accuracy": 0.9270695447921753,
|
| 1193 |
+
"step": 148
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 4.806451612903226,
|
| 1197 |
+
"grad_norm": 0.1444924771785736,
|
| 1198 |
+
"learning_rate": 9.180725568338043e-10,
|
| 1199 |
+
"loss": 0.3102,
|
| 1200 |
+
"mean_token_accuracy": 0.9180653691291809,
|
| 1201 |
+
"step": 149
|
| 1202 |
+
},
|
| 1203 |
+
{
|
| 1204 |
+
"epoch": 4.838709677419355,
|
| 1205 |
+
"grad_norm": 0.15837052464485168,
|
| 1206 |
+
"learning_rate": 6.37849069761176e-10,
|
| 1207 |
+
"loss": 0.328,
|
| 1208 |
+
"mean_token_accuracy": 0.9148820638656616,
|
| 1209 |
+
"step": 150
|
| 1210 |
+
},
|
| 1211 |
+
{
|
| 1212 |
+
"epoch": 4.870967741935484,
|
| 1213 |
+
"grad_norm": 0.14807510375976562,
|
| 1214 |
+
"learning_rate": 4.083798592444898e-10,
|
| 1215 |
+
"loss": 0.3087,
|
| 1216 |
+
"mean_token_accuracy": 0.9184255003929138,
|
| 1217 |
+
"step": 151
|
| 1218 |
+
},
|
| 1219 |
+
{
|
| 1220 |
+
"epoch": 4.903225806451613,
|
| 1221 |
+
"grad_norm": 0.15544399619102478,
|
| 1222 |
+
"learning_rate": 2.2978213836400973e-10,
|
| 1223 |
+
"loss": 0.2967,
|
| 1224 |
+
"mean_token_accuracy": 0.9193904995918274,
|
| 1225 |
+
"step": 152
|
| 1226 |
+
},
|
| 1227 |
+
{
|
| 1228 |
+
"epoch": 4.935483870967742,
|
| 1229 |
+
"grad_norm": 0.18428154289722443,
|
| 1230 |
+
"learning_rate": 1.0214713499706595e-10,
|
| 1231 |
+
"loss": 0.3394,
|
| 1232 |
+
"mean_token_accuracy": 0.9154376983642578,
|
| 1233 |
+
"step": 153
|
| 1234 |
+
},
|
| 1235 |
+
{
|
| 1236 |
+
"epoch": 4.967741935483871,
|
| 1237 |
+
"grad_norm": 0.13660480082035065,
|
| 1238 |
+
"learning_rate": 2.554004521881925e-11,
|
| 1239 |
+
"loss": 0.2831,
|
| 1240 |
+
"mean_token_accuracy": 0.9276332855224609,
|
| 1241 |
+
"step": 154
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 5.0,
|
| 1245 |
+
"grad_norm": 420961255424.0,
|
| 1246 |
+
"learning_rate": 0.0,
|
| 1247 |
+
"loss": 0.3237,
|
| 1248 |
+
"mean_token_accuracy": 0.9185903668403625,
|
| 1249 |
+
"step": 155
|
| 1250 |
+
},
|
| 1251 |
+
{
|
| 1252 |
+
"epoch": 5.0,
|
| 1253 |
+
"step": 155,
|
| 1254 |
+
"total_flos": 1.846643377789993e+17,
|
| 1255 |
+
"train_loss": 0.30200490220900506,
|
| 1256 |
+
"train_runtime": 1464.9032,
|
| 1257 |
+
"train_samples_per_second": 6.741,
|
| 1258 |
+
"train_steps_per_second": 0.106
|
| 1259 |
+
}
|
| 1260 |
+
],
|
| 1261 |
+
"logging_steps": 1,
|
| 1262 |
+
"max_steps": 155,
|
| 1263 |
+
"num_input_tokens_seen": 0,
|
| 1264 |
+
"num_train_epochs": 5,
|
| 1265 |
+
"save_steps": 1,
|
| 1266 |
+
"stateful_callbacks": {
|
| 1267 |
+
"TrainerControl": {
|
| 1268 |
+
"args": {
|
| 1269 |
+
"should_epoch_stop": false,
|
| 1270 |
+
"should_evaluate": false,
|
| 1271 |
+
"should_log": false,
|
| 1272 |
+
"should_save": true,
|
| 1273 |
+
"should_training_stop": true
|
| 1274 |
+
},
|
| 1275 |
+
"attributes": {}
|
| 1276 |
+
}
|
| 1277 |
+
},
|
| 1278 |
+
"total_flos": 1.846643377789993e+17,
|
| 1279 |
+
"train_batch_size": 8,
|
| 1280 |
+
"trial_name": null,
|
| 1281 |
+
"trial_params": null
|
| 1282 |
+
}
|