Model save
Browse files- README.md +58 -0
- all_results.json +8 -0
- train_results.json +8 -0
- trainer_state.json +2422 -0
README.md
ADDED
|
@@ -0,0 +1,58 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
base_model: Qwen/Qwen2.5-1.5B-Instruct
|
| 3 |
+
library_name: transformers
|
| 4 |
+
model_name: Qwen2.5-1.5B-Open-R1-Distill-lora
|
| 5 |
+
tags:
|
| 6 |
+
- generated_from_trainer
|
| 7 |
+
- trl
|
| 8 |
+
- sft
|
| 9 |
+
licence: license
|
| 10 |
+
---
|
| 11 |
+
|
| 12 |
+
# Model Card for Qwen2.5-1.5B-Open-R1-Distill-lora
|
| 13 |
+
|
| 14 |
+
This model is a fine-tuned version of [Qwen/Qwen2.5-1.5B-Instruct](https://huggingface.co/Qwen/Qwen2.5-1.5B-Instruct).
|
| 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="T1anyu/Qwen2.5-1.5B-Open-R1-Distill-lora", 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/tianyufan0504_/huggingface/runs/uny3r3cv)
|
| 31 |
+
|
| 32 |
+
|
| 33 |
+
This model was trained with SFT.
|
| 34 |
+
|
| 35 |
+
### Framework versions
|
| 36 |
+
|
| 37 |
+
- TRL: 0.15.0
|
| 38 |
+
- Transformers: 4.50.0.dev0
|
| 39 |
+
- Pytorch: 2.5.1
|
| 40 |
+
- Datasets: 3.3.1
|
| 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 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"total_flos": 2.629944131882844e+18,
|
| 3 |
+
"train_loss": 0.6986723100437837,
|
| 4 |
+
"train_runtime": 41334.8641,
|
| 5 |
+
"train_samples": 93733,
|
| 6 |
+
"train_samples_per_second": 0.481,
|
| 7 |
+
"train_steps_per_second": 0.041
|
| 8 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"total_flos": 2.629944131882844e+18,
|
| 3 |
+
"train_loss": 0.6986723100437837,
|
| 4 |
+
"train_runtime": 41334.8641,
|
| 5 |
+
"train_samples": 93733,
|
| 6 |
+
"train_samples_per_second": 0.481,
|
| 7 |
+
"train_steps_per_second": 0.041
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2422 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 50.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 1700,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.14705882352941177,
|
| 13 |
+
"grad_norm": 0.14238034188747406,
|
| 14 |
+
"learning_rate": 2.9411764705882355e-06,
|
| 15 |
+
"loss": 0.8446,
|
| 16 |
+
"step": 5
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.29411764705882354,
|
| 20 |
+
"grad_norm": 0.14489340782165527,
|
| 21 |
+
"learning_rate": 5.882352941176471e-06,
|
| 22 |
+
"loss": 0.8425,
|
| 23 |
+
"step": 10
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.4411764705882353,
|
| 27 |
+
"grad_norm": 0.1551039218902588,
|
| 28 |
+
"learning_rate": 8.823529411764707e-06,
|
| 29 |
+
"loss": 0.8918,
|
| 30 |
+
"step": 15
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.5882352941176471,
|
| 34 |
+
"grad_norm": 0.15667259693145752,
|
| 35 |
+
"learning_rate": 1.1764705882352942e-05,
|
| 36 |
+
"loss": 0.8473,
|
| 37 |
+
"step": 20
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.7352941176470589,
|
| 41 |
+
"grad_norm": 0.14174659550189972,
|
| 42 |
+
"learning_rate": 1.4705882352941177e-05,
|
| 43 |
+
"loss": 0.8519,
|
| 44 |
+
"step": 25
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.8823529411764706,
|
| 48 |
+
"grad_norm": 0.15220147371292114,
|
| 49 |
+
"learning_rate": 1.7647058823529414e-05,
|
| 50 |
+
"loss": 0.8201,
|
| 51 |
+
"step": 30
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 1.0294117647058822,
|
| 55 |
+
"grad_norm": 0.14088748395442963,
|
| 56 |
+
"learning_rate": 2.058823529411765e-05,
|
| 57 |
+
"loss": 0.8345,
|
| 58 |
+
"step": 35
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 1.1764705882352942,
|
| 62 |
+
"grad_norm": 0.1412580907344818,
|
| 63 |
+
"learning_rate": 2.3529411764705884e-05,
|
| 64 |
+
"loss": 0.8549,
|
| 65 |
+
"step": 40
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 1.3235294117647058,
|
| 69 |
+
"grad_norm": 0.12632407248020172,
|
| 70 |
+
"learning_rate": 2.647058823529412e-05,
|
| 71 |
+
"loss": 0.8431,
|
| 72 |
+
"step": 45
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 1.4705882352941178,
|
| 76 |
+
"grad_norm": 0.11696045845746994,
|
| 77 |
+
"learning_rate": 2.9411764705882354e-05,
|
| 78 |
+
"loss": 0.8157,
|
| 79 |
+
"step": 50
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 1.6176470588235294,
|
| 83 |
+
"grad_norm": 0.1272636502981186,
|
| 84 |
+
"learning_rate": 3.235294117647059e-05,
|
| 85 |
+
"loss": 0.8236,
|
| 86 |
+
"step": 55
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 1.7647058823529411,
|
| 90 |
+
"grad_norm": 0.10813312977552414,
|
| 91 |
+
"learning_rate": 3.529411764705883e-05,
|
| 92 |
+
"loss": 0.8019,
|
| 93 |
+
"step": 60
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 1.9117647058823528,
|
| 97 |
+
"grad_norm": 0.11896699666976929,
|
| 98 |
+
"learning_rate": 3.8235294117647055e-05,
|
| 99 |
+
"loss": 0.8185,
|
| 100 |
+
"step": 65
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 2.0588235294117645,
|
| 104 |
+
"grad_norm": 0.10265181213617325,
|
| 105 |
+
"learning_rate": 4.11764705882353e-05,
|
| 106 |
+
"loss": 0.8044,
|
| 107 |
+
"step": 70
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 2.2058823529411766,
|
| 111 |
+
"grad_norm": 0.10106126219034195,
|
| 112 |
+
"learning_rate": 4.411764705882353e-05,
|
| 113 |
+
"loss": 0.7858,
|
| 114 |
+
"step": 75
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 2.3529411764705883,
|
| 118 |
+
"grad_norm": 0.10131911188364029,
|
| 119 |
+
"learning_rate": 4.705882352941177e-05,
|
| 120 |
+
"loss": 0.8084,
|
| 121 |
+
"step": 80
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 2.5,
|
| 125 |
+
"grad_norm": 0.09996718913316727,
|
| 126 |
+
"learning_rate": 5e-05,
|
| 127 |
+
"loss": 0.814,
|
| 128 |
+
"step": 85
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 2.6470588235294117,
|
| 132 |
+
"grad_norm": 0.0959576964378357,
|
| 133 |
+
"learning_rate": 4.999893574965545e-05,
|
| 134 |
+
"loss": 0.8105,
|
| 135 |
+
"step": 90
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 2.7941176470588234,
|
| 139 |
+
"grad_norm": 0.11985825002193451,
|
| 140 |
+
"learning_rate": 4.9995743099299886e-05,
|
| 141 |
+
"loss": 0.7813,
|
| 142 |
+
"step": 95
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 2.9411764705882355,
|
| 146 |
+
"grad_norm": 0.09690634161233902,
|
| 147 |
+
"learning_rate": 4.9990422350958156e-05,
|
| 148 |
+
"loss": 0.7917,
|
| 149 |
+
"step": 100
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 3.088235294117647,
|
| 153 |
+
"grad_norm": 0.11959923803806305,
|
| 154 |
+
"learning_rate": 4.99829740079732e-05,
|
| 155 |
+
"loss": 0.7531,
|
| 156 |
+
"step": 105
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 3.235294117647059,
|
| 160 |
+
"grad_norm": 0.09686623513698578,
|
| 161 |
+
"learning_rate": 4.99733987749585e-05,
|
| 162 |
+
"loss": 0.783,
|
| 163 |
+
"step": 110
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 3.3823529411764706,
|
| 167 |
+
"grad_norm": 0.0896935760974884,
|
| 168 |
+
"learning_rate": 4.996169755773138e-05,
|
| 169 |
+
"loss": 0.7506,
|
| 170 |
+
"step": 115
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 3.5294117647058822,
|
| 174 |
+
"grad_norm": 0.09192899614572525,
|
| 175 |
+
"learning_rate": 4.9947871463227374e-05,
|
| 176 |
+
"loss": 0.7597,
|
| 177 |
+
"step": 120
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 3.6764705882352944,
|
| 181 |
+
"grad_norm": 0.09252317994832993,
|
| 182 |
+
"learning_rate": 4.993192179939542e-05,
|
| 183 |
+
"loss": 0.767,
|
| 184 |
+
"step": 125
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 3.8235294117647056,
|
| 188 |
+
"grad_norm": 0.09957632422447205,
|
| 189 |
+
"learning_rate": 4.991385007507422e-05,
|
| 190 |
+
"loss": 0.8006,
|
| 191 |
+
"step": 130
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 3.9705882352941178,
|
| 195 |
+
"grad_norm": 0.12217137217521667,
|
| 196 |
+
"learning_rate": 4.989365799984943e-05,
|
| 197 |
+
"loss": 0.7513,
|
| 198 |
+
"step": 135
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 4.117647058823529,
|
| 202 |
+
"grad_norm": 0.09010059386491776,
|
| 203 |
+
"learning_rate": 4.9871347483892006e-05,
|
| 204 |
+
"loss": 0.7531,
|
| 205 |
+
"step": 140
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 4.264705882352941,
|
| 209 |
+
"grad_norm": 0.09149183332920074,
|
| 210 |
+
"learning_rate": 4.984692063777743e-05,
|
| 211 |
+
"loss": 0.7518,
|
| 212 |
+
"step": 145
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 4.411764705882353,
|
| 216 |
+
"grad_norm": 0.09295113384723663,
|
| 217 |
+
"learning_rate": 4.9820379772286095e-05,
|
| 218 |
+
"loss": 0.7665,
|
| 219 |
+
"step": 150
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 4.5588235294117645,
|
| 223 |
+
"grad_norm": 0.09854214638471603,
|
| 224 |
+
"learning_rate": 4.979172739818469e-05,
|
| 225 |
+
"loss": 0.7739,
|
| 226 |
+
"step": 155
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 4.705882352941177,
|
| 230 |
+
"grad_norm": 0.09157629311084747,
|
| 231 |
+
"learning_rate": 4.9760966225988675e-05,
|
| 232 |
+
"loss": 0.7522,
|
| 233 |
+
"step": 160
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 4.852941176470588,
|
| 237 |
+
"grad_norm": 0.10833761096000671,
|
| 238 |
+
"learning_rate": 4.9728099165705895e-05,
|
| 239 |
+
"loss": 0.7605,
|
| 240 |
+
"step": 165
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 5.0,
|
| 244 |
+
"grad_norm": 0.09924355149269104,
|
| 245 |
+
"learning_rate": 4.9693129326561254e-05,
|
| 246 |
+
"loss": 0.7153,
|
| 247 |
+
"step": 170
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 5.147058823529412,
|
| 251 |
+
"grad_norm": 0.09313185513019562,
|
| 252 |
+
"learning_rate": 4.9656060016702606e-05,
|
| 253 |
+
"loss": 0.7494,
|
| 254 |
+
"step": 175
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 5.294117647058823,
|
| 258 |
+
"grad_norm": 0.11171400547027588,
|
| 259 |
+
"learning_rate": 4.961689474288779e-05,
|
| 260 |
+
"loss": 0.733,
|
| 261 |
+
"step": 180
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 5.4411764705882355,
|
| 265 |
+
"grad_norm": 0.09828388690948486,
|
| 266 |
+
"learning_rate": 4.957563721015293e-05,
|
| 267 |
+
"loss": 0.7663,
|
| 268 |
+
"step": 185
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 5.588235294117647,
|
| 272 |
+
"grad_norm": 0.09972433745861053,
|
| 273 |
+
"learning_rate": 4.953229132146186e-05,
|
| 274 |
+
"loss": 0.7576,
|
| 275 |
+
"step": 190
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 5.735294117647059,
|
| 279 |
+
"grad_norm": 0.11432339251041412,
|
| 280 |
+
"learning_rate": 4.948686117733699e-05,
|
| 281 |
+
"loss": 0.7379,
|
| 282 |
+
"step": 195
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 5.882352941176471,
|
| 286 |
+
"grad_norm": 0.10343588888645172,
|
| 287 |
+
"learning_rate": 4.9439351075471346e-05,
|
| 288 |
+
"loss": 0.7066,
|
| 289 |
+
"step": 200
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 6.029411764705882,
|
| 293 |
+
"grad_norm": 0.0964265912771225,
|
| 294 |
+
"learning_rate": 4.9389765510322026e-05,
|
| 295 |
+
"loss": 0.7322,
|
| 296 |
+
"step": 205
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 6.176470588235294,
|
| 300 |
+
"grad_norm": 0.11457476019859314,
|
| 301 |
+
"learning_rate": 4.9338109172685006e-05,
|
| 302 |
+
"loss": 0.742,
|
| 303 |
+
"step": 210
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 6.323529411764706,
|
| 307 |
+
"grad_norm": 0.10812544822692871,
|
| 308 |
+
"learning_rate": 4.92843869492514e-05,
|
| 309 |
+
"loss": 0.7572,
|
| 310 |
+
"step": 215
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 6.470588235294118,
|
| 314 |
+
"grad_norm": 0.1057206243276596,
|
| 315 |
+
"learning_rate": 4.9228603922145206e-05,
|
| 316 |
+
"loss": 0.7342,
|
| 317 |
+
"step": 220
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 6.617647058823529,
|
| 321 |
+
"grad_norm": 0.11412467062473297,
|
| 322 |
+
"learning_rate": 4.917076536844248e-05,
|
| 323 |
+
"loss": 0.7331,
|
| 324 |
+
"step": 225
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 6.764705882352941,
|
| 328 |
+
"grad_norm": 0.11059483885765076,
|
| 329 |
+
"learning_rate": 4.9110876759672184e-05,
|
| 330 |
+
"loss": 0.718,
|
| 331 |
+
"step": 230
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 6.911764705882353,
|
| 335 |
+
"grad_norm": 0.10819140076637268,
|
| 336 |
+
"learning_rate": 4.9048943761298544e-05,
|
| 337 |
+
"loss": 0.7153,
|
| 338 |
+
"step": 235
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 7.0588235294117645,
|
| 342 |
+
"grad_norm": 0.11002287268638611,
|
| 343 |
+
"learning_rate": 4.89849722321851e-05,
|
| 344 |
+
"loss": 0.7201,
|
| 345 |
+
"step": 240
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 7.205882352941177,
|
| 349 |
+
"grad_norm": 0.11289830505847931,
|
| 350 |
+
"learning_rate": 4.891896822404046e-05,
|
| 351 |
+
"loss": 0.7261,
|
| 352 |
+
"step": 245
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 7.352941176470588,
|
| 356 |
+
"grad_norm": 0.12590822577476501,
|
| 357 |
+
"learning_rate": 4.885093798084583e-05,
|
| 358 |
+
"loss": 0.7329,
|
| 359 |
+
"step": 250
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 7.5,
|
| 363 |
+
"grad_norm": 0.10964758694171906,
|
| 364 |
+
"learning_rate": 4.878088793826428e-05,
|
| 365 |
+
"loss": 0.7413,
|
| 366 |
+
"step": 255
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 7.647058823529412,
|
| 370 |
+
"grad_norm": 0.10680090636014938,
|
| 371 |
+
"learning_rate": 4.8708824723031995e-05,
|
| 372 |
+
"loss": 0.7174,
|
| 373 |
+
"step": 260
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 7.794117647058823,
|
| 377 |
+
"grad_norm": 0.10832036286592484,
|
| 378 |
+
"learning_rate": 4.8634755152331355e-05,
|
| 379 |
+
"loss": 0.7345,
|
| 380 |
+
"step": 265
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 7.9411764705882355,
|
| 384 |
+
"grad_norm": 0.10789214819669724,
|
| 385 |
+
"learning_rate": 4.8558686233145996e-05,
|
| 386 |
+
"loss": 0.7213,
|
| 387 |
+
"step": 270
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 8.088235294117647,
|
| 391 |
+
"grad_norm": 0.11145245283842087,
|
| 392 |
+
"learning_rate": 4.8480625161598e-05,
|
| 393 |
+
"loss": 0.7184,
|
| 394 |
+
"step": 275
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 8.235294117647058,
|
| 398 |
+
"grad_norm": 0.12131233513355255,
|
| 399 |
+
"learning_rate": 4.840057932226715e-05,
|
| 400 |
+
"loss": 0.737,
|
| 401 |
+
"step": 280
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 8.382352941176471,
|
| 405 |
+
"grad_norm": 0.11897268146276474,
|
| 406 |
+
"learning_rate": 4.831855628749228e-05,
|
| 407 |
+
"loss": 0.7254,
|
| 408 |
+
"step": 285
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 8.529411764705882,
|
| 412 |
+
"grad_norm": 0.1130763441324234,
|
| 413 |
+
"learning_rate": 4.823456381665501e-05,
|
| 414 |
+
"loss": 0.7213,
|
| 415 |
+
"step": 290
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 8.676470588235293,
|
| 419 |
+
"grad_norm": 0.11835259944200516,
|
| 420 |
+
"learning_rate": 4.8148609855445624e-05,
|
| 421 |
+
"loss": 0.7102,
|
| 422 |
+
"step": 295
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 8.823529411764707,
|
| 426 |
+
"grad_norm": 0.12309901416301727,
|
| 427 |
+
"learning_rate": 4.806070253511151e-05,
|
| 428 |
+
"loss": 0.7227,
|
| 429 |
+
"step": 300
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 8.970588235294118,
|
| 433 |
+
"grad_norm": 0.11361519992351532,
|
| 434 |
+
"learning_rate": 4.797085017168787e-05,
|
| 435 |
+
"loss": 0.7125,
|
| 436 |
+
"step": 305
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 9.117647058823529,
|
| 440 |
+
"grad_norm": 0.13154913485050201,
|
| 441 |
+
"learning_rate": 4.7879061265211e-05,
|
| 442 |
+
"loss": 0.7293,
|
| 443 |
+
"step": 310
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 9.264705882352942,
|
| 447 |
+
"grad_norm": 0.12245271354913712,
|
| 448 |
+
"learning_rate": 4.778534449891428e-05,
|
| 449 |
+
"loss": 0.7216,
|
| 450 |
+
"step": 315
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 9.411764705882353,
|
| 454 |
+
"grad_norm": 0.10899204015731812,
|
| 455 |
+
"learning_rate": 4.768970873840669e-05,
|
| 456 |
+
"loss": 0.706,
|
| 457 |
+
"step": 320
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 9.558823529411764,
|
| 461 |
+
"grad_norm": 0.1145118996500969,
|
| 462 |
+
"learning_rate": 4.75921630308341e-05,
|
| 463 |
+
"loss": 0.7039,
|
| 464 |
+
"step": 325
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 9.705882352941176,
|
| 468 |
+
"grad_norm": 0.12111522257328033,
|
| 469 |
+
"learning_rate": 4.749271660402341e-05,
|
| 470 |
+
"loss": 0.7359,
|
| 471 |
+
"step": 330
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 9.852941176470589,
|
| 475 |
+
"grad_norm": 0.11270228773355484,
|
| 476 |
+
"learning_rate": 4.739137886560966e-05,
|
| 477 |
+
"loss": 0.7006,
|
| 478 |
+
"step": 335
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 10.0,
|
| 482 |
+
"grad_norm": 0.10989291220903397,
|
| 483 |
+
"learning_rate": 4.7288159402146e-05,
|
| 484 |
+
"loss": 0.7123,
|
| 485 |
+
"step": 340
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 10.147058823529411,
|
| 489 |
+
"grad_norm": 0.11979430168867111,
|
| 490 |
+
"learning_rate": 4.7183067978196855e-05,
|
| 491 |
+
"loss": 0.7213,
|
| 492 |
+
"step": 345
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 10.294117647058824,
|
| 496 |
+
"grad_norm": 0.11735141277313232,
|
| 497 |
+
"learning_rate": 4.707611453541412e-05,
|
| 498 |
+
"loss": 0.7061,
|
| 499 |
+
"step": 350
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 10.441176470588236,
|
| 503 |
+
"grad_norm": 0.12181384861469269,
|
| 504 |
+
"learning_rate": 4.696730919159677e-05,
|
| 505 |
+
"loss": 0.6962,
|
| 506 |
+
"step": 355
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 10.588235294117647,
|
| 510 |
+
"grad_norm": 0.11275137960910797,
|
| 511 |
+
"learning_rate": 4.6856662239733666e-05,
|
| 512 |
+
"loss": 0.7467,
|
| 513 |
+
"step": 360
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 10.735294117647058,
|
| 517 |
+
"grad_norm": 0.13028523325920105,
|
| 518 |
+
"learning_rate": 4.674418414702985e-05,
|
| 519 |
+
"loss": 0.7047,
|
| 520 |
+
"step": 365
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 10.882352941176471,
|
| 524 |
+
"grad_norm": 0.12034178525209427,
|
| 525 |
+
"learning_rate": 4.662988555391632e-05,
|
| 526 |
+
"loss": 0.7061,
|
| 527 |
+
"step": 370
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 11.029411764705882,
|
| 531 |
+
"grad_norm": 0.11595606803894043,
|
| 532 |
+
"learning_rate": 4.6513777273043495e-05,
|
| 533 |
+
"loss": 0.7023,
|
| 534 |
+
"step": 375
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 11.176470588235293,
|
| 538 |
+
"grad_norm": 0.11920719593763351,
|
| 539 |
+
"learning_rate": 4.63958702882583e-05,
|
| 540 |
+
"loss": 0.6886,
|
| 541 |
+
"step": 380
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 11.323529411764707,
|
| 545 |
+
"grad_norm": 0.12535597383975983,
|
| 546 |
+
"learning_rate": 4.6276175753565105e-05,
|
| 547 |
+
"loss": 0.7209,
|
| 548 |
+
"step": 385
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 11.470588235294118,
|
| 552 |
+
"grad_norm": 0.12857039272785187,
|
| 553 |
+
"learning_rate": 4.615470499207056e-05,
|
| 554 |
+
"loss": 0.7018,
|
| 555 |
+
"step": 390
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 11.617647058823529,
|
| 559 |
+
"grad_norm": 0.13531994819641113,
|
| 560 |
+
"learning_rate": 4.6031469494912416e-05,
|
| 561 |
+
"loss": 0.7145,
|
| 562 |
+
"step": 395
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 11.764705882352942,
|
| 566 |
+
"grad_norm": 0.10658453404903412,
|
| 567 |
+
"learning_rate": 4.59064809201725e-05,
|
| 568 |
+
"loss": 0.723,
|
| 569 |
+
"step": 400
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 11.911764705882353,
|
| 573 |
+
"grad_norm": 0.12011521309614182,
|
| 574 |
+
"learning_rate": 4.5779751091773774e-05,
|
| 575 |
+
"loss": 0.7011,
|
| 576 |
+
"step": 405
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 12.058823529411764,
|
| 580 |
+
"grad_norm": 0.11478458344936371,
|
| 581 |
+
"learning_rate": 4.5651291998361926e-05,
|
| 582 |
+
"loss": 0.7117,
|
| 583 |
+
"step": 410
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 12.205882352941176,
|
| 587 |
+
"grad_norm": 0.1333089917898178,
|
| 588 |
+
"learning_rate": 4.55211157921711e-05,
|
| 589 |
+
"loss": 0.7148,
|
| 590 |
+
"step": 415
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 12.352941176470589,
|
| 594 |
+
"grad_norm": 0.11957768350839615,
|
| 595 |
+
"learning_rate": 4.538923478787439e-05,
|
| 596 |
+
"loss": 0.7049,
|
| 597 |
+
"step": 420
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 12.5,
|
| 601 |
+
"grad_norm": 0.11590797454118729,
|
| 602 |
+
"learning_rate": 4.5255661461418854e-05,
|
| 603 |
+
"loss": 0.6797,
|
| 604 |
+
"step": 425
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 12.647058823529411,
|
| 608 |
+
"grad_norm": 0.12927260994911194,
|
| 609 |
+
"learning_rate": 4.5120408448845264e-05,
|
| 610 |
+
"loss": 0.7126,
|
| 611 |
+
"step": 430
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 12.794117647058824,
|
| 615 |
+
"grad_norm": 0.13119827210903168,
|
| 616 |
+
"learning_rate": 4.4983488545092753e-05,
|
| 617 |
+
"loss": 0.7082,
|
| 618 |
+
"step": 435
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 12.941176470588236,
|
| 622 |
+
"grad_norm": 0.1294083297252655,
|
| 623 |
+
"learning_rate": 4.4844914702788386e-05,
|
| 624 |
+
"loss": 0.699,
|
| 625 |
+
"step": 440
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 13.088235294117647,
|
| 629 |
+
"grad_norm": 0.11839265376329422,
|
| 630 |
+
"learning_rate": 4.470470003102192e-05,
|
| 631 |
+
"loss": 0.71,
|
| 632 |
+
"step": 445
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 13.235294117647058,
|
| 636 |
+
"grad_norm": 0.12290767580270767,
|
| 637 |
+
"learning_rate": 4.456285779410558e-05,
|
| 638 |
+
"loss": 0.7058,
|
| 639 |
+
"step": 450
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 13.382352941176471,
|
| 643 |
+
"grad_norm": 0.12060663849115372,
|
| 644 |
+
"learning_rate": 4.4419401410319334e-05,
|
| 645 |
+
"loss": 0.6744,
|
| 646 |
+
"step": 455
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 13.529411764705882,
|
| 650 |
+
"grad_norm": 0.12197393923997879,
|
| 651 |
+
"learning_rate": 4.427434445064148e-05,
|
| 652 |
+
"loss": 0.6919,
|
| 653 |
+
"step": 460
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 13.676470588235293,
|
| 657 |
+
"grad_norm": 0.12659871578216553,
|
| 658 |
+
"learning_rate": 4.4127700637464834e-05,
|
| 659 |
+
"loss": 0.7102,
|
| 660 |
+
"step": 465
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 13.823529411764707,
|
| 664 |
+
"grad_norm": 0.12616273760795593,
|
| 665 |
+
"learning_rate": 4.3979483843298624e-05,
|
| 666 |
+
"loss": 0.6924,
|
| 667 |
+
"step": 470
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 13.970588235294118,
|
| 671 |
+
"grad_norm": 0.1336318999528885,
|
| 672 |
+
"learning_rate": 4.382970808945612e-05,
|
| 673 |
+
"loss": 0.7248,
|
| 674 |
+
"step": 475
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 14.117647058823529,
|
| 678 |
+
"grad_norm": 0.12975798547267914,
|
| 679 |
+
"learning_rate": 4.367838754472821e-05,
|
| 680 |
+
"loss": 0.7266,
|
| 681 |
+
"step": 480
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 14.264705882352942,
|
| 685 |
+
"grad_norm": 0.1267329454421997,
|
| 686 |
+
"learning_rate": 4.3525536524043076e-05,
|
| 687 |
+
"loss": 0.7028,
|
| 688 |
+
"step": 485
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 14.411764705882353,
|
| 692 |
+
"grad_norm": 0.12307338416576385,
|
| 693 |
+
"learning_rate": 4.337116948711195e-05,
|
| 694 |
+
"loss": 0.7052,
|
| 695 |
+
"step": 490
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 14.558823529411764,
|
| 699 |
+
"grad_norm": 0.14381247758865356,
|
| 700 |
+
"learning_rate": 4.3215301037061244e-05,
|
| 701 |
+
"loss": 0.6947,
|
| 702 |
+
"step": 495
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 14.705882352941176,
|
| 706 |
+
"grad_norm": 0.11929916590452194,
|
| 707 |
+
"learning_rate": 4.305794591905113e-05,
|
| 708 |
+
"loss": 0.691,
|
| 709 |
+
"step": 500
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 14.852941176470589,
|
| 713 |
+
"grad_norm": 0.12451142817735672,
|
| 714 |
+
"learning_rate": 4.289911901888056e-05,
|
| 715 |
+
"loss": 0.6859,
|
| 716 |
+
"step": 505
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 15.0,
|
| 720 |
+
"grad_norm": 0.12542015314102173,
|
| 721 |
+
"learning_rate": 4.2738835361579175e-05,
|
| 722 |
+
"loss": 0.7139,
|
| 723 |
+
"step": 510
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 15.147058823529411,
|
| 727 |
+
"grad_norm": 0.11630310118198395,
|
| 728 |
+
"learning_rate": 4.257711010998586e-05,
|
| 729 |
+
"loss": 0.705,
|
| 730 |
+
"step": 515
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 15.294117647058824,
|
| 734 |
+
"grad_norm": 0.13719907402992249,
|
| 735 |
+
"learning_rate": 4.241395856331437e-05,
|
| 736 |
+
"loss": 0.7001,
|
| 737 |
+
"step": 520
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 15.441176470588236,
|
| 741 |
+
"grad_norm": 0.13168473541736603,
|
| 742 |
+
"learning_rate": 4.224939615570602e-05,
|
| 743 |
+
"loss": 0.7047,
|
| 744 |
+
"step": 525
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 15.588235294117647,
|
| 748 |
+
"grad_norm": 0.11908990889787674,
|
| 749 |
+
"learning_rate": 4.2083438454769606e-05,
|
| 750 |
+
"loss": 0.7086,
|
| 751 |
+
"step": 530
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 15.735294117647058,
|
| 755 |
+
"grad_norm": 0.12613283097743988,
|
| 756 |
+
"learning_rate": 4.1916101160108715e-05,
|
| 757 |
+
"loss": 0.6911,
|
| 758 |
+
"step": 535
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 15.882352941176471,
|
| 762 |
+
"grad_norm": 0.12481500208377838,
|
| 763 |
+
"learning_rate": 4.174740010183656e-05,
|
| 764 |
+
"loss": 0.6845,
|
| 765 |
+
"step": 540
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 16.029411764705884,
|
| 769 |
+
"grad_norm": 0.1311793029308319,
|
| 770 |
+
"learning_rate": 4.15773512390784e-05,
|
| 771 |
+
"loss": 0.6976,
|
| 772 |
+
"step": 545
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 16.176470588235293,
|
| 776 |
+
"grad_norm": 0.11820737272500992,
|
| 777 |
+
"learning_rate": 4.140597065846188e-05,
|
| 778 |
+
"loss": 0.7101,
|
| 779 |
+
"step": 550
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 16.323529411764707,
|
| 783 |
+
"grad_norm": 0.12259554117918015,
|
| 784 |
+
"learning_rate": 4.123327457259517e-05,
|
| 785 |
+
"loss": 0.6973,
|
| 786 |
+
"step": 555
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 16.470588235294116,
|
| 790 |
+
"grad_norm": 0.12584823369979858,
|
| 791 |
+
"learning_rate": 4.105927931853327e-05,
|
| 792 |
+
"loss": 0.6903,
|
| 793 |
+
"step": 560
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 16.61764705882353,
|
| 797 |
+
"grad_norm": 0.13863661885261536,
|
| 798 |
+
"learning_rate": 4.088400135623256e-05,
|
| 799 |
+
"loss": 0.6726,
|
| 800 |
+
"step": 565
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 16.764705882352942,
|
| 804 |
+
"grad_norm": 0.13399486243724823,
|
| 805 |
+
"learning_rate": 4.070745726699363e-05,
|
| 806 |
+
"loss": 0.6977,
|
| 807 |
+
"step": 570
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 16.91176470588235,
|
| 811 |
+
"grad_norm": 0.1234726756811142,
|
| 812 |
+
"learning_rate": 4.0529663751892734e-05,
|
| 813 |
+
"loss": 0.6907,
|
| 814 |
+
"step": 575
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 17.058823529411764,
|
| 818 |
+
"grad_norm": 0.11935710906982422,
|
| 819 |
+
"learning_rate": 4.035063763020185e-05,
|
| 820 |
+
"loss": 0.7128,
|
| 821 |
+
"step": 580
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 17.205882352941178,
|
| 825 |
+
"grad_norm": 0.12929606437683105,
|
| 826 |
+
"learning_rate": 4.017039583779756e-05,
|
| 827 |
+
"loss": 0.7106,
|
| 828 |
+
"step": 585
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 17.352941176470587,
|
| 832 |
+
"grad_norm": 0.1248982772231102,
|
| 833 |
+
"learning_rate": 3.9988955425558965e-05,
|
| 834 |
+
"loss": 0.6897,
|
| 835 |
+
"step": 590
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 17.5,
|
| 839 |
+
"grad_norm": 0.1281740367412567,
|
| 840 |
+
"learning_rate": 3.980633355775461e-05,
|
| 841 |
+
"loss": 0.6871,
|
| 842 |
+
"step": 595
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 17.647058823529413,
|
| 846 |
+
"grad_norm": 0.12564094364643097,
|
| 847 |
+
"learning_rate": 3.962254751041877e-05,
|
| 848 |
+
"loss": 0.7008,
|
| 849 |
+
"step": 600
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 17.794117647058822,
|
| 853 |
+
"grad_norm": 0.1336313784122467,
|
| 854 |
+
"learning_rate": 3.943761466971717e-05,
|
| 855 |
+
"loss": 0.6851,
|
| 856 |
+
"step": 605
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 17.941176470588236,
|
| 860 |
+
"grad_norm": 0.1354963481426239,
|
| 861 |
+
"learning_rate": 3.9251552530302206e-05,
|
| 862 |
+
"loss": 0.6951,
|
| 863 |
+
"step": 610
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 18.08823529411765,
|
| 867 |
+
"grad_norm": 0.15230301022529602,
|
| 868 |
+
"learning_rate": 3.906437869365795e-05,
|
| 869 |
+
"loss": 0.693,
|
| 870 |
+
"step": 615
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 18.235294117647058,
|
| 874 |
+
"grad_norm": 0.1296202540397644,
|
| 875 |
+
"learning_rate": 3.887611086643508e-05,
|
| 876 |
+
"loss": 0.6874,
|
| 877 |
+
"step": 620
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 18.38235294117647,
|
| 881 |
+
"grad_norm": 0.13370412588119507,
|
| 882 |
+
"learning_rate": 3.8686766858775843e-05,
|
| 883 |
+
"loss": 0.7085,
|
| 884 |
+
"step": 625
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 18.529411764705884,
|
| 888 |
+
"grad_norm": 0.13132105767726898,
|
| 889 |
+
"learning_rate": 3.849636458262913e-05,
|
| 890 |
+
"loss": 0.7037,
|
| 891 |
+
"step": 630
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 18.676470588235293,
|
| 895 |
+
"grad_norm": 0.13370752334594727,
|
| 896 |
+
"learning_rate": 3.830492205005612e-05,
|
| 897 |
+
"loss": 0.679,
|
| 898 |
+
"step": 635
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 18.823529411764707,
|
| 902 |
+
"grad_norm": 0.12619805335998535,
|
| 903 |
+
"learning_rate": 3.811245737152624e-05,
|
| 904 |
+
"loss": 0.6846,
|
| 905 |
+
"step": 640
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 18.970588235294116,
|
| 909 |
+
"grad_norm": 0.13043031096458435,
|
| 910 |
+
"learning_rate": 3.7918988754203985e-05,
|
| 911 |
+
"loss": 0.6729,
|
| 912 |
+
"step": 645
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 19.11764705882353,
|
| 916 |
+
"grad_norm": 0.132464200258255,
|
| 917 |
+
"learning_rate": 3.772453450022649e-05,
|
| 918 |
+
"loss": 0.7112,
|
| 919 |
+
"step": 650
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 19.264705882352942,
|
| 923 |
+
"grad_norm": 0.1268807351589203,
|
| 924 |
+
"learning_rate": 3.752911300497212e-05,
|
| 925 |
+
"loss": 0.6804,
|
| 926 |
+
"step": 655
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 19.41176470588235,
|
| 930 |
+
"grad_norm": 0.14288154244422913,
|
| 931 |
+
"learning_rate": 3.73327427553203e-05,
|
| 932 |
+
"loss": 0.6867,
|
| 933 |
+
"step": 660
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 19.558823529411764,
|
| 937 |
+
"grad_norm": 0.14849698543548584,
|
| 938 |
+
"learning_rate": 3.7135442327902695e-05,
|
| 939 |
+
"loss": 0.694,
|
| 940 |
+
"step": 665
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 19.705882352941178,
|
| 944 |
+
"grad_norm": 0.12607896327972412,
|
| 945 |
+
"learning_rate": 3.6937230387345746e-05,
|
| 946 |
+
"loss": 0.6873,
|
| 947 |
+
"step": 670
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 19.852941176470587,
|
| 951 |
+
"grad_norm": 0.12860074639320374,
|
| 952 |
+
"learning_rate": 3.673812568450513e-05,
|
| 953 |
+
"loss": 0.6942,
|
| 954 |
+
"step": 675
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 20.0,
|
| 958 |
+
"grad_norm": 0.12468240410089493,
|
| 959 |
+
"learning_rate": 3.6538147054691817e-05,
|
| 960 |
+
"loss": 0.6844,
|
| 961 |
+
"step": 680
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 20.147058823529413,
|
| 965 |
+
"grad_norm": 0.1367003470659256,
|
| 966 |
+
"learning_rate": 3.6337313415890315e-05,
|
| 967 |
+
"loss": 0.7005,
|
| 968 |
+
"step": 685
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 20.294117647058822,
|
| 972 |
+
"grad_norm": 0.13072577118873596,
|
| 973 |
+
"learning_rate": 3.6135643766969e-05,
|
| 974 |
+
"loss": 0.671,
|
| 975 |
+
"step": 690
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 20.441176470588236,
|
| 979 |
+
"grad_norm": 0.1326008439064026,
|
| 980 |
+
"learning_rate": 3.593315718588286e-05,
|
| 981 |
+
"loss": 0.6727,
|
| 982 |
+
"step": 695
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 20.58823529411765,
|
| 986 |
+
"grad_norm": 0.1257023960351944,
|
| 987 |
+
"learning_rate": 3.572987282786864e-05,
|
| 988 |
+
"loss": 0.7073,
|
| 989 |
+
"step": 700
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 20.735294117647058,
|
| 993 |
+
"grad_norm": 0.14335250854492188,
|
| 994 |
+
"learning_rate": 3.552580992363285e-05,
|
| 995 |
+
"loss": 0.6821,
|
| 996 |
+
"step": 705
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 20.88235294117647,
|
| 1000 |
+
"grad_norm": 0.13271793723106384,
|
| 1001 |
+
"learning_rate": 3.5320987777532465e-05,
|
| 1002 |
+
"loss": 0.6959,
|
| 1003 |
+
"step": 710
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 21.029411764705884,
|
| 1007 |
+
"grad_norm": 0.12265238165855408,
|
| 1008 |
+
"learning_rate": 3.5115425765748793e-05,
|
| 1009 |
+
"loss": 0.6767,
|
| 1010 |
+
"step": 715
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 21.176470588235293,
|
| 1014 |
+
"grad_norm": 0.13558083772659302,
|
| 1015 |
+
"learning_rate": 3.4909143334454454e-05,
|
| 1016 |
+
"loss": 0.6859,
|
| 1017 |
+
"step": 720
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 21.323529411764707,
|
| 1021 |
+
"grad_norm": 0.1432723104953766,
|
| 1022 |
+
"learning_rate": 3.4702159997973747e-05,
|
| 1023 |
+
"loss": 0.6921,
|
| 1024 |
+
"step": 725
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 21.470588235294116,
|
| 1028 |
+
"grad_norm": 0.13662855327129364,
|
| 1029 |
+
"learning_rate": 3.449449533693664e-05,
|
| 1030 |
+
"loss": 0.7063,
|
| 1031 |
+
"step": 730
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 21.61764705882353,
|
| 1035 |
+
"grad_norm": 0.1422967165708542,
|
| 1036 |
+
"learning_rate": 3.428616899642645e-05,
|
| 1037 |
+
"loss": 0.6987,
|
| 1038 |
+
"step": 735
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 21.764705882352942,
|
| 1042 |
+
"grad_norm": 0.1233050599694252,
|
| 1043 |
+
"learning_rate": 3.4077200684121345e-05,
|
| 1044 |
+
"loss": 0.6831,
|
| 1045 |
+
"step": 740
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 21.91176470588235,
|
| 1049 |
+
"grad_norm": 0.13728494942188263,
|
| 1050 |
+
"learning_rate": 3.3867610168430084e-05,
|
| 1051 |
+
"loss": 0.6873,
|
| 1052 |
+
"step": 745
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 22.058823529411764,
|
| 1056 |
+
"grad_norm": 0.1322290301322937,
|
| 1057 |
+
"learning_rate": 3.365741727662187e-05,
|
| 1058 |
+
"loss": 0.651,
|
| 1059 |
+
"step": 750
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 22.205882352941178,
|
| 1063 |
+
"grad_norm": 0.13796144723892212,
|
| 1064 |
+
"learning_rate": 3.3446641892950696e-05,
|
| 1065 |
+
"loss": 0.671,
|
| 1066 |
+
"step": 755
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 22.352941176470587,
|
| 1070 |
+
"grad_norm": 0.13293515145778656,
|
| 1071 |
+
"learning_rate": 3.3235303956774324e-05,
|
| 1072 |
+
"loss": 0.7056,
|
| 1073 |
+
"step": 760
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 22.5,
|
| 1077 |
+
"grad_norm": 0.13630028069019318,
|
| 1078 |
+
"learning_rate": 3.3023423460667985e-05,
|
| 1079 |
+
"loss": 0.6866,
|
| 1080 |
+
"step": 765
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 22.647058823529413,
|
| 1084 |
+
"grad_norm": 0.1360238939523697,
|
| 1085 |
+
"learning_rate": 3.281102044853309e-05,
|
| 1086 |
+
"loss": 0.6991,
|
| 1087 |
+
"step": 770
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 22.794117647058822,
|
| 1091 |
+
"grad_norm": 0.13794392347335815,
|
| 1092 |
+
"learning_rate": 3.2598115013701114e-05,
|
| 1093 |
+
"loss": 0.6959,
|
| 1094 |
+
"step": 775
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 22.941176470588236,
|
| 1098 |
+
"grad_norm": 0.1369139850139618,
|
| 1099 |
+
"learning_rate": 3.2384727297032705e-05,
|
| 1100 |
+
"loss": 0.6657,
|
| 1101 |
+
"step": 780
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 23.08823529411765,
|
| 1105 |
+
"grad_norm": 0.12718403339385986,
|
| 1106 |
+
"learning_rate": 3.217087748501237e-05,
|
| 1107 |
+
"loss": 0.6733,
|
| 1108 |
+
"step": 785
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 23.235294117647058,
|
| 1112 |
+
"grad_norm": 0.13353672623634338,
|
| 1113 |
+
"learning_rate": 3.1956585807838914e-05,
|
| 1114 |
+
"loss": 0.6774,
|
| 1115 |
+
"step": 790
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 23.38235294117647,
|
| 1119 |
+
"grad_norm": 0.13364404439926147,
|
| 1120 |
+
"learning_rate": 3.1741872537511535e-05,
|
| 1121 |
+
"loss": 0.6752,
|
| 1122 |
+
"step": 795
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 23.529411764705884,
|
| 1126 |
+
"grad_norm": 0.14464542269706726,
|
| 1127 |
+
"learning_rate": 3.152675798591219e-05,
|
| 1128 |
+
"loss": 0.6667,
|
| 1129 |
+
"step": 800
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 23.676470588235293,
|
| 1133 |
+
"grad_norm": 0.13043712079524994,
|
| 1134 |
+
"learning_rate": 3.131126250288405e-05,
|
| 1135 |
+
"loss": 0.6924,
|
| 1136 |
+
"step": 805
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 23.823529411764707,
|
| 1140 |
+
"grad_norm": 0.12341820448637009,
|
| 1141 |
+
"learning_rate": 3.109540647430641e-05,
|
| 1142 |
+
"loss": 0.6969,
|
| 1143 |
+
"step": 810
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 23.970588235294116,
|
| 1147 |
+
"grad_norm": 0.14009004831314087,
|
| 1148 |
+
"learning_rate": 3.087921032016619e-05,
|
| 1149 |
+
"loss": 0.6947,
|
| 1150 |
+
"step": 815
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 24.11764705882353,
|
| 1154 |
+
"grad_norm": 0.13528534770011902,
|
| 1155 |
+
"learning_rate": 3.066269449262618e-05,
|
| 1156 |
+
"loss": 0.6833,
|
| 1157 |
+
"step": 820
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 24.264705882352942,
|
| 1161 |
+
"grad_norm": 0.1405653953552246,
|
| 1162 |
+
"learning_rate": 3.04458794740903e-05,
|
| 1163 |
+
"loss": 0.6919,
|
| 1164 |
+
"step": 825
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 24.41176470588235,
|
| 1168 |
+
"grad_norm": 0.14450417459011078,
|
| 1169 |
+
"learning_rate": 3.0228785775265943e-05,
|
| 1170 |
+
"loss": 0.7085,
|
| 1171 |
+
"step": 830
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 24.558823529411764,
|
| 1175 |
+
"grad_norm": 0.1257210224866867,
|
| 1176 |
+
"learning_rate": 3.001143393322368e-05,
|
| 1177 |
+
"loss": 0.7022,
|
| 1178 |
+
"step": 835
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 24.705882352941178,
|
| 1182 |
+
"grad_norm": 0.14650067687034607,
|
| 1183 |
+
"learning_rate": 2.9793844509454417e-05,
|
| 1184 |
+
"loss": 0.6559,
|
| 1185 |
+
"step": 840
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 24.852941176470587,
|
| 1189 |
+
"grad_norm": 0.1673348844051361,
|
| 1190 |
+
"learning_rate": 2.9576038087924297e-05,
|
| 1191 |
+
"loss": 0.6628,
|
| 1192 |
+
"step": 845
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 25.0,
|
| 1196 |
+
"grad_norm": 0.13868844509124756,
|
| 1197 |
+
"learning_rate": 2.9358035273127483e-05,
|
| 1198 |
+
"loss": 0.6761,
|
| 1199 |
+
"step": 850
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 25.147058823529413,
|
| 1203 |
+
"grad_norm": 0.12067105621099472,
|
| 1204 |
+
"learning_rate": 2.9139856688136917e-05,
|
| 1205 |
+
"loss": 0.6735,
|
| 1206 |
+
"step": 855
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 25.294117647058822,
|
| 1210 |
+
"grad_norm": 0.1306021362543106,
|
| 1211 |
+
"learning_rate": 2.8921522972653437e-05,
|
| 1212 |
+
"loss": 0.6711,
|
| 1213 |
+
"step": 860
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 25.441176470588236,
|
| 1217 |
+
"grad_norm": 0.13525615632534027,
|
| 1218 |
+
"learning_rate": 2.8703054781053194e-05,
|
| 1219 |
+
"loss": 0.6723,
|
| 1220 |
+
"step": 865
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 25.58823529411765,
|
| 1224 |
+
"grad_norm": 0.1306258738040924,
|
| 1225 |
+
"learning_rate": 2.8484472780433828e-05,
|
| 1226 |
+
"loss": 0.6922,
|
| 1227 |
+
"step": 870
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 25.735294117647058,
|
| 1231 |
+
"grad_norm": 0.14182746410369873,
|
| 1232 |
+
"learning_rate": 2.8265797648659283e-05,
|
| 1233 |
+
"loss": 0.6911,
|
| 1234 |
+
"step": 875
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 25.88235294117647,
|
| 1238 |
+
"grad_norm": 0.13599254190921783,
|
| 1239 |
+
"learning_rate": 2.8047050072403713e-05,
|
| 1240 |
+
"loss": 0.6891,
|
| 1241 |
+
"step": 880
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 26.029411764705884,
|
| 1245 |
+
"grad_norm": 0.1291087120771408,
|
| 1246 |
+
"learning_rate": 2.7828250745194544e-05,
|
| 1247 |
+
"loss": 0.6971,
|
| 1248 |
+
"step": 885
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 26.176470588235293,
|
| 1252 |
+
"grad_norm": 0.11979696899652481,
|
| 1253 |
+
"learning_rate": 2.7609420365454823e-05,
|
| 1254 |
+
"loss": 0.6921,
|
| 1255 |
+
"step": 890
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 26.323529411764707,
|
| 1259 |
+
"grad_norm": 0.1369645744562149,
|
| 1260 |
+
"learning_rate": 2.7390579634545182e-05,
|
| 1261 |
+
"loss": 0.667,
|
| 1262 |
+
"step": 895
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 26.470588235294116,
|
| 1266 |
+
"grad_norm": 0.1354684978723526,
|
| 1267 |
+
"learning_rate": 2.7171749254805458e-05,
|
| 1268 |
+
"loss": 0.6918,
|
| 1269 |
+
"step": 900
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 26.61764705882353,
|
| 1273 |
+
"grad_norm": 0.1434841752052307,
|
| 1274 |
+
"learning_rate": 2.6952949927596295e-05,
|
| 1275 |
+
"loss": 0.6961,
|
| 1276 |
+
"step": 905
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 26.764705882352942,
|
| 1280 |
+
"grad_norm": 0.13030685484409332,
|
| 1281 |
+
"learning_rate": 2.6734202351340726e-05,
|
| 1282 |
+
"loss": 0.6742,
|
| 1283 |
+
"step": 910
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 26.91176470588235,
|
| 1287 |
+
"grad_norm": 0.1375734657049179,
|
| 1288 |
+
"learning_rate": 2.651552721956617e-05,
|
| 1289 |
+
"loss": 0.66,
|
| 1290 |
+
"step": 915
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 27.058823529411764,
|
| 1294 |
+
"grad_norm": 0.1508912891149521,
|
| 1295 |
+
"learning_rate": 2.6296945218946804e-05,
|
| 1296 |
+
"loss": 0.6928,
|
| 1297 |
+
"step": 920
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 27.205882352941178,
|
| 1301 |
+
"grad_norm": 0.13976359367370605,
|
| 1302 |
+
"learning_rate": 2.6078477027346572e-05,
|
| 1303 |
+
"loss": 0.6916,
|
| 1304 |
+
"step": 925
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 27.352941176470587,
|
| 1308 |
+
"grad_norm": 0.13399522006511688,
|
| 1309 |
+
"learning_rate": 2.586014331186309e-05,
|
| 1310 |
+
"loss": 0.6617,
|
| 1311 |
+
"step": 930
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 27.5,
|
| 1315 |
+
"grad_norm": 0.1378486156463623,
|
| 1316 |
+
"learning_rate": 2.5641964726872526e-05,
|
| 1317 |
+
"loss": 0.6779,
|
| 1318 |
+
"step": 935
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 27.647058823529413,
|
| 1322 |
+
"grad_norm": 0.1410367488861084,
|
| 1323 |
+
"learning_rate": 2.5423961912075712e-05,
|
| 1324 |
+
"loss": 0.6951,
|
| 1325 |
+
"step": 940
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 27.794117647058822,
|
| 1329 |
+
"grad_norm": 0.1448415368795395,
|
| 1330 |
+
"learning_rate": 2.5206155490545585e-05,
|
| 1331 |
+
"loss": 0.6958,
|
| 1332 |
+
"step": 945
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 27.941176470588236,
|
| 1336 |
+
"grad_norm": 0.1381085067987442,
|
| 1337 |
+
"learning_rate": 2.4988566066776327e-05,
|
| 1338 |
+
"loss": 0.6629,
|
| 1339 |
+
"step": 950
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 28.08823529411765,
|
| 1343 |
+
"grad_norm": 0.14611601829528809,
|
| 1344 |
+
"learning_rate": 2.4771214224734056e-05,
|
| 1345 |
+
"loss": 0.6642,
|
| 1346 |
+
"step": 955
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 28.235294117647058,
|
| 1350 |
+
"grad_norm": 0.13046316802501678,
|
| 1351 |
+
"learning_rate": 2.4554120525909703e-05,
|
| 1352 |
+
"loss": 0.6554,
|
| 1353 |
+
"step": 960
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 28.38235294117647,
|
| 1357 |
+
"grad_norm": 0.1373993307352066,
|
| 1358 |
+
"learning_rate": 2.4337305507373832e-05,
|
| 1359 |
+
"loss": 0.6791,
|
| 1360 |
+
"step": 965
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 28.529411764705884,
|
| 1364 |
+
"grad_norm": 0.140591099858284,
|
| 1365 |
+
"learning_rate": 2.4120789679833815e-05,
|
| 1366 |
+
"loss": 0.6729,
|
| 1367 |
+
"step": 970
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 28.676470588235293,
|
| 1371 |
+
"grad_norm": 0.1307932734489441,
|
| 1372 |
+
"learning_rate": 2.3904593525693593e-05,
|
| 1373 |
+
"loss": 0.6887,
|
| 1374 |
+
"step": 975
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 28.823529411764707,
|
| 1378 |
+
"grad_norm": 0.13051795959472656,
|
| 1379 |
+
"learning_rate": 2.3688737497115953e-05,
|
| 1380 |
+
"loss": 0.6823,
|
| 1381 |
+
"step": 980
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 28.970588235294116,
|
| 1385 |
+
"grad_norm": 0.12720821797847748,
|
| 1386 |
+
"learning_rate": 2.3473242014087814e-05,
|
| 1387 |
+
"loss": 0.7063,
|
| 1388 |
+
"step": 985
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 29.11764705882353,
|
| 1392 |
+
"grad_norm": 0.137127086520195,
|
| 1393 |
+
"learning_rate": 2.3258127462488467e-05,
|
| 1394 |
+
"loss": 0.6744,
|
| 1395 |
+
"step": 990
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 29.264705882352942,
|
| 1399 |
+
"grad_norm": 0.13432453572750092,
|
| 1400 |
+
"learning_rate": 2.30434141921611e-05,
|
| 1401 |
+
"loss": 0.68,
|
| 1402 |
+
"step": 995
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 29.41176470588235,
|
| 1406 |
+
"grad_norm": 0.14380089938640594,
|
| 1407 |
+
"learning_rate": 2.2829122514987634e-05,
|
| 1408 |
+
"loss": 0.6808,
|
| 1409 |
+
"step": 1000
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 29.558823529411764,
|
| 1413 |
+
"grad_norm": 0.12999729812145233,
|
| 1414 |
+
"learning_rate": 2.2615272702967304e-05,
|
| 1415 |
+
"loss": 0.6963,
|
| 1416 |
+
"step": 1005
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 29.705882352941178,
|
| 1420 |
+
"grad_norm": 0.13407659530639648,
|
| 1421 |
+
"learning_rate": 2.2401884986298892e-05,
|
| 1422 |
+
"loss": 0.6729,
|
| 1423 |
+
"step": 1010
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 29.852941176470587,
|
| 1427 |
+
"grad_norm": 0.13908743858337402,
|
| 1428 |
+
"learning_rate": 2.2188979551466916e-05,
|
| 1429 |
+
"loss": 0.6766,
|
| 1430 |
+
"step": 1015
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 30.0,
|
| 1434 |
+
"grad_norm": 0.13133063912391663,
|
| 1435 |
+
"learning_rate": 2.1976576539332024e-05,
|
| 1436 |
+
"loss": 0.664,
|
| 1437 |
+
"step": 1020
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 30.147058823529413,
|
| 1441 |
+
"grad_norm": 0.15663307905197144,
|
| 1442 |
+
"learning_rate": 2.1764696043225685e-05,
|
| 1443 |
+
"loss": 0.7082,
|
| 1444 |
+
"step": 1025
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 30.294117647058822,
|
| 1448 |
+
"grad_norm": 0.13505025207996368,
|
| 1449 |
+
"learning_rate": 2.155335810704931e-05,
|
| 1450 |
+
"loss": 0.6463,
|
| 1451 |
+
"step": 1030
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 30.441176470588236,
|
| 1455 |
+
"grad_norm": 0.1344403475522995,
|
| 1456 |
+
"learning_rate": 2.134258272337814e-05,
|
| 1457 |
+
"loss": 0.6753,
|
| 1458 |
+
"step": 1035
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 30.58823529411765,
|
| 1462 |
+
"grad_norm": 0.14067409932613373,
|
| 1463 |
+
"learning_rate": 2.1132389831569915e-05,
|
| 1464 |
+
"loss": 0.6715,
|
| 1465 |
+
"step": 1040
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 30.735294117647058,
|
| 1469 |
+
"grad_norm": 0.13444367051124573,
|
| 1470 |
+
"learning_rate": 2.092279931587866e-05,
|
| 1471 |
+
"loss": 0.6838,
|
| 1472 |
+
"step": 1045
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 30.88235294117647,
|
| 1476 |
+
"grad_norm": 0.13275469839572906,
|
| 1477 |
+
"learning_rate": 2.0713831003573564e-05,
|
| 1478 |
+
"loss": 0.6842,
|
| 1479 |
+
"step": 1050
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 31.029411764705884,
|
| 1483 |
+
"grad_norm": 0.12724100053310394,
|
| 1484 |
+
"learning_rate": 2.0505504663063364e-05,
|
| 1485 |
+
"loss": 0.6745,
|
| 1486 |
+
"step": 1055
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 31.176470588235293,
|
| 1490 |
+
"grad_norm": 0.12783651053905487,
|
| 1491 |
+
"learning_rate": 2.029784000202627e-05,
|
| 1492 |
+
"loss": 0.6839,
|
| 1493 |
+
"step": 1060
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 31.323529411764707,
|
| 1497 |
+
"grad_norm": 0.13505741953849792,
|
| 1498 |
+
"learning_rate": 2.0090856665545554e-05,
|
| 1499 |
+
"loss": 0.6577,
|
| 1500 |
+
"step": 1065
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 31.470588235294116,
|
| 1504 |
+
"grad_norm": 0.14324721693992615,
|
| 1505 |
+
"learning_rate": 1.98845742342512e-05,
|
| 1506 |
+
"loss": 0.6786,
|
| 1507 |
+
"step": 1070
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 31.61764705882353,
|
| 1511 |
+
"grad_norm": 0.14350536465644836,
|
| 1512 |
+
"learning_rate": 1.967901222246754e-05,
|
| 1513 |
+
"loss": 0.6715,
|
| 1514 |
+
"step": 1075
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 31.764705882352942,
|
| 1518 |
+
"grad_norm": 0.12864898145198822,
|
| 1519 |
+
"learning_rate": 1.947419007636716e-05,
|
| 1520 |
+
"loss": 0.6901,
|
| 1521 |
+
"step": 1080
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 31.91176470588235,
|
| 1525 |
+
"grad_norm": 0.13163405656814575,
|
| 1526 |
+
"learning_rate": 1.9270127172131363e-05,
|
| 1527 |
+
"loss": 0.6767,
|
| 1528 |
+
"step": 1085
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 32.05882352941177,
|
| 1532 |
+
"grad_norm": 0.13823044300079346,
|
| 1533 |
+
"learning_rate": 1.906684281411715e-05,
|
| 1534 |
+
"loss": 0.6888,
|
| 1535 |
+
"step": 1090
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 32.205882352941174,
|
| 1539 |
+
"grad_norm": 0.13260214030742645,
|
| 1540 |
+
"learning_rate": 1.8864356233031e-05,
|
| 1541 |
+
"loss": 0.6899,
|
| 1542 |
+
"step": 1095
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 32.35294117647059,
|
| 1546 |
+
"grad_norm": 0.13542212545871735,
|
| 1547 |
+
"learning_rate": 1.866268658410969e-05,
|
| 1548 |
+
"loss": 0.6604,
|
| 1549 |
+
"step": 1100
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 32.5,
|
| 1553 |
+
"grad_norm": 0.14194779098033905,
|
| 1554 |
+
"learning_rate": 1.8461852945308196e-05,
|
| 1555 |
+
"loss": 0.6538,
|
| 1556 |
+
"step": 1105
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 32.64705882352941,
|
| 1560 |
+
"grad_norm": 0.13551092147827148,
|
| 1561 |
+
"learning_rate": 1.8261874315494874e-05,
|
| 1562 |
+
"loss": 0.6851,
|
| 1563 |
+
"step": 1110
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 32.794117647058826,
|
| 1567 |
+
"grad_norm": 0.13539521396160126,
|
| 1568 |
+
"learning_rate": 1.806276961265425e-05,
|
| 1569 |
+
"loss": 0.6731,
|
| 1570 |
+
"step": 1115
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 32.94117647058823,
|
| 1574 |
+
"grad_norm": 0.14235951006412506,
|
| 1575 |
+
"learning_rate": 1.786455767209732e-05,
|
| 1576 |
+
"loss": 0.6798,
|
| 1577 |
+
"step": 1120
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 33.088235294117645,
|
| 1581 |
+
"grad_norm": 0.12894190847873688,
|
| 1582 |
+
"learning_rate": 1.7667257244679702e-05,
|
| 1583 |
+
"loss": 0.6815,
|
| 1584 |
+
"step": 1125
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 33.23529411764706,
|
| 1588 |
+
"grad_norm": 0.13332705199718475,
|
| 1589 |
+
"learning_rate": 1.747088699502789e-05,
|
| 1590 |
+
"loss": 0.6709,
|
| 1591 |
+
"step": 1130
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 33.38235294117647,
|
| 1595 |
+
"grad_norm": 0.13527055084705353,
|
| 1596 |
+
"learning_rate": 1.727546549977352e-05,
|
| 1597 |
+
"loss": 0.689,
|
| 1598 |
+
"step": 1135
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 33.529411764705884,
|
| 1602 |
+
"grad_norm": 0.13612490892410278,
|
| 1603 |
+
"learning_rate": 1.7081011245796013e-05,
|
| 1604 |
+
"loss": 0.6744,
|
| 1605 |
+
"step": 1140
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 33.6764705882353,
|
| 1609 |
+
"grad_norm": 0.13099683821201324,
|
| 1610 |
+
"learning_rate": 1.6887542628473763e-05,
|
| 1611 |
+
"loss": 0.6871,
|
| 1612 |
+
"step": 1145
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 33.8235294117647,
|
| 1616 |
+
"grad_norm": 0.13698424398899078,
|
| 1617 |
+
"learning_rate": 1.6695077949943892e-05,
|
| 1618 |
+
"loss": 0.6852,
|
| 1619 |
+
"step": 1150
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 33.970588235294116,
|
| 1623 |
+
"grad_norm": 0.13121846318244934,
|
| 1624 |
+
"learning_rate": 1.6503635417370882e-05,
|
| 1625 |
+
"loss": 0.6529,
|
| 1626 |
+
"step": 1155
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 34.11764705882353,
|
| 1630 |
+
"grad_norm": 0.1369757056236267,
|
| 1631 |
+
"learning_rate": 1.6313233141224165e-05,
|
| 1632 |
+
"loss": 0.6855,
|
| 1633 |
+
"step": 1160
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 34.26470588235294,
|
| 1637 |
+
"grad_norm": 0.13654442131519318,
|
| 1638 |
+
"learning_rate": 1.612388913356493e-05,
|
| 1639 |
+
"loss": 0.6596,
|
| 1640 |
+
"step": 1165
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 34.411764705882355,
|
| 1644 |
+
"grad_norm": 0.136439248919487,
|
| 1645 |
+
"learning_rate": 1.5935621306342057e-05,
|
| 1646 |
+
"loss": 0.6843,
|
| 1647 |
+
"step": 1170
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 34.55882352941177,
|
| 1651 |
+
"grad_norm": 0.1410278081893921,
|
| 1652 |
+
"learning_rate": 1.5748447469697803e-05,
|
| 1653 |
+
"loss": 0.6786,
|
| 1654 |
+
"step": 1175
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 34.705882352941174,
|
| 1658 |
+
"grad_norm": 0.16095899045467377,
|
| 1659 |
+
"learning_rate": 1.556238533028283e-05,
|
| 1660 |
+
"loss": 0.6563,
|
| 1661 |
+
"step": 1180
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 34.85294117647059,
|
| 1665 |
+
"grad_norm": 0.13262508809566498,
|
| 1666 |
+
"learning_rate": 1.5377452489581234e-05,
|
| 1667 |
+
"loss": 0.6888,
|
| 1668 |
+
"step": 1185
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 35.0,
|
| 1672 |
+
"grad_norm": 0.13472063839435577,
|
| 1673 |
+
"learning_rate": 1.5193666442245402e-05,
|
| 1674 |
+
"loss": 0.681,
|
| 1675 |
+
"step": 1190
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 35.14705882352941,
|
| 1679 |
+
"grad_norm": 0.13549183309078217,
|
| 1680 |
+
"learning_rate": 1.5011044574441036e-05,
|
| 1681 |
+
"loss": 0.6755,
|
| 1682 |
+
"step": 1195
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 35.294117647058826,
|
| 1686 |
+
"grad_norm": 0.1411600261926651,
|
| 1687 |
+
"learning_rate": 1.4829604162202442e-05,
|
| 1688 |
+
"loss": 0.7007,
|
| 1689 |
+
"step": 1200
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 35.44117647058823,
|
| 1693 |
+
"grad_norm": 0.14127956330776215,
|
| 1694 |
+
"learning_rate": 1.4649362369798152e-05,
|
| 1695 |
+
"loss": 0.6551,
|
| 1696 |
+
"step": 1205
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 35.588235294117645,
|
| 1700 |
+
"grad_norm": 0.13209925591945648,
|
| 1701 |
+
"learning_rate": 1.4470336248107266e-05,
|
| 1702 |
+
"loss": 0.6762,
|
| 1703 |
+
"step": 1210
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 35.73529411764706,
|
| 1707 |
+
"grad_norm": 0.12888824939727783,
|
| 1708 |
+
"learning_rate": 1.4292542733006372e-05,
|
| 1709 |
+
"loss": 0.6775,
|
| 1710 |
+
"step": 1215
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 35.88235294117647,
|
| 1714 |
+
"grad_norm": 0.14431186020374298,
|
| 1715 |
+
"learning_rate": 1.4115998643767447e-05,
|
| 1716 |
+
"loss": 0.6654,
|
| 1717 |
+
"step": 1220
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 36.029411764705884,
|
| 1721 |
+
"grad_norm": 0.12955108284950256,
|
| 1722 |
+
"learning_rate": 1.3940720681466734e-05,
|
| 1723 |
+
"loss": 0.6807,
|
| 1724 |
+
"step": 1225
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 36.1764705882353,
|
| 1728 |
+
"grad_norm": 0.13727155327796936,
|
| 1729 |
+
"learning_rate": 1.3766725427404843e-05,
|
| 1730 |
+
"loss": 0.6925,
|
| 1731 |
+
"step": 1230
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 36.3235294117647,
|
| 1735 |
+
"grad_norm": 0.13375459611415863,
|
| 1736 |
+
"learning_rate": 1.3594029341538128e-05,
|
| 1737 |
+
"loss": 0.6884,
|
| 1738 |
+
"step": 1235
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 36.470588235294116,
|
| 1742 |
+
"grad_norm": 0.13129761815071106,
|
| 1743 |
+
"learning_rate": 1.34226487609216e-05,
|
| 1744 |
+
"loss": 0.6868,
|
| 1745 |
+
"step": 1240
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 36.61764705882353,
|
| 1749 |
+
"grad_norm": 0.1358431726694107,
|
| 1750 |
+
"learning_rate": 1.3252599898163454e-05,
|
| 1751 |
+
"loss": 0.6538,
|
| 1752 |
+
"step": 1245
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 36.76470588235294,
|
| 1756 |
+
"grad_norm": 0.14378570020198822,
|
| 1757 |
+
"learning_rate": 1.3083898839891284e-05,
|
| 1758 |
+
"loss": 0.6457,
|
| 1759 |
+
"step": 1250
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 36.911764705882355,
|
| 1763 |
+
"grad_norm": 0.14497900009155273,
|
| 1764 |
+
"learning_rate": 1.29165615452304e-05,
|
| 1765 |
+
"loss": 0.6746,
|
| 1766 |
+
"step": 1255
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 37.05882352941177,
|
| 1770 |
+
"grad_norm": 0.13402092456817627,
|
| 1771 |
+
"learning_rate": 1.275060384429398e-05,
|
| 1772 |
+
"loss": 0.6721,
|
| 1773 |
+
"step": 1260
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 37.205882352941174,
|
| 1777 |
+
"grad_norm": 0.13633766770362854,
|
| 1778 |
+
"learning_rate": 1.258604143668563e-05,
|
| 1779 |
+
"loss": 0.6724,
|
| 1780 |
+
"step": 1265
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 37.35294117647059,
|
| 1784 |
+
"grad_norm": 0.13490943610668182,
|
| 1785 |
+
"learning_rate": 1.2422889890014143e-05,
|
| 1786 |
+
"loss": 0.6578,
|
| 1787 |
+
"step": 1270
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 37.5,
|
| 1791 |
+
"grad_norm": 0.1326485574245453,
|
| 1792 |
+
"learning_rate": 1.2261164638420832e-05,
|
| 1793 |
+
"loss": 0.6664,
|
| 1794 |
+
"step": 1275
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 37.64705882352941,
|
| 1798 |
+
"grad_norm": 0.14504876732826233,
|
| 1799 |
+
"learning_rate": 1.2100880981119447e-05,
|
| 1800 |
+
"loss": 0.6856,
|
| 1801 |
+
"step": 1280
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 37.794117647058826,
|
| 1805 |
+
"grad_norm": 0.13638907670974731,
|
| 1806 |
+
"learning_rate": 1.1942054080948878e-05,
|
| 1807 |
+
"loss": 0.6842,
|
| 1808 |
+
"step": 1285
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 37.94117647058823,
|
| 1812 |
+
"grad_norm": 0.16528142988681793,
|
| 1813 |
+
"learning_rate": 1.1784698962938763e-05,
|
| 1814 |
+
"loss": 0.6759,
|
| 1815 |
+
"step": 1290
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 38.088235294117645,
|
| 1819 |
+
"grad_norm": 0.15061551332473755,
|
| 1820 |
+
"learning_rate": 1.1628830512888057e-05,
|
| 1821 |
+
"loss": 0.6899,
|
| 1822 |
+
"step": 1295
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 38.23529411764706,
|
| 1826 |
+
"grad_norm": 0.13696105778217316,
|
| 1827 |
+
"learning_rate": 1.1474463475956926e-05,
|
| 1828 |
+
"loss": 0.6624,
|
| 1829 |
+
"step": 1300
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 38.38235294117647,
|
| 1833 |
+
"grad_norm": 0.12491544336080551,
|
| 1834 |
+
"learning_rate": 1.1321612455271793e-05,
|
| 1835 |
+
"loss": 0.6725,
|
| 1836 |
+
"step": 1305
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 38.529411764705884,
|
| 1840 |
+
"grad_norm": 0.13985736668109894,
|
| 1841 |
+
"learning_rate": 1.117029191054389e-05,
|
| 1842 |
+
"loss": 0.6942,
|
| 1843 |
+
"step": 1310
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 38.6764705882353,
|
| 1847 |
+
"grad_norm": 0.14015409350395203,
|
| 1848 |
+
"learning_rate": 1.1020516156701383e-05,
|
| 1849 |
+
"loss": 0.6759,
|
| 1850 |
+
"step": 1315
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 38.8235294117647,
|
| 1854 |
+
"grad_norm": 0.14540641009807587,
|
| 1855 |
+
"learning_rate": 1.0872299362535173e-05,
|
| 1856 |
+
"loss": 0.6645,
|
| 1857 |
+
"step": 1320
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 38.970588235294116,
|
| 1861 |
+
"grad_norm": 0.1425599455833435,
|
| 1862 |
+
"learning_rate": 1.0725655549358532e-05,
|
| 1863 |
+
"loss": 0.6711,
|
| 1864 |
+
"step": 1325
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 39.11764705882353,
|
| 1868 |
+
"grad_norm": 0.13927870988845825,
|
| 1869 |
+
"learning_rate": 1.0580598589680664e-05,
|
| 1870 |
+
"loss": 0.6956,
|
| 1871 |
+
"step": 1330
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 39.26470588235294,
|
| 1875 |
+
"grad_norm": 0.13224616646766663,
|
| 1876 |
+
"learning_rate": 1.0437142205894418e-05,
|
| 1877 |
+
"loss": 0.6868,
|
| 1878 |
+
"step": 1335
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 39.411764705882355,
|
| 1882 |
+
"grad_norm": 0.13682135939598083,
|
| 1883 |
+
"learning_rate": 1.029529996897808e-05,
|
| 1884 |
+
"loss": 0.6735,
|
| 1885 |
+
"step": 1340
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 39.55882352941177,
|
| 1889 |
+
"grad_norm": 0.1319390833377838,
|
| 1890 |
+
"learning_rate": 1.0155085297211618e-05,
|
| 1891 |
+
"loss": 0.6513,
|
| 1892 |
+
"step": 1345
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 39.705882352941174,
|
| 1896 |
+
"grad_norm": 0.1452108919620514,
|
| 1897 |
+
"learning_rate": 1.001651145490726e-05,
|
| 1898 |
+
"loss": 0.6772,
|
| 1899 |
+
"step": 1350
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 39.85294117647059,
|
| 1903 |
+
"grad_norm": 0.14989398419857025,
|
| 1904 |
+
"learning_rate": 9.87959155115474e-06,
|
| 1905 |
+
"loss": 0.6633,
|
| 1906 |
+
"step": 1355
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 40.0,
|
| 1910 |
+
"grad_norm": 0.14220058917999268,
|
| 1911 |
+
"learning_rate": 9.744338538581147e-06,
|
| 1912 |
+
"loss": 0.6778,
|
| 1913 |
+
"step": 1360
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 40.14705882352941,
|
| 1917 |
+
"grad_norm": 0.15334346890449524,
|
| 1918 |
+
"learning_rate": 9.610765212125607e-06,
|
| 1919 |
+
"loss": 0.6775,
|
| 1920 |
+
"step": 1365
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 40.294117647058826,
|
| 1924 |
+
"grad_norm": 0.1356540322303772,
|
| 1925 |
+
"learning_rate": 9.478884207828912e-06,
|
| 1926 |
+
"loss": 0.6513,
|
| 1927 |
+
"step": 1370
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 40.44117647058823,
|
| 1931 |
+
"grad_norm": 0.14519663155078888,
|
| 1932 |
+
"learning_rate": 9.34870800163808e-06,
|
| 1933 |
+
"loss": 0.6847,
|
| 1934 |
+
"step": 1375
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 40.588235294117645,
|
| 1938 |
+
"grad_norm": 0.13579830527305603,
|
| 1939 |
+
"learning_rate": 9.220248908226224e-06,
|
| 1940 |
+
"loss": 0.6661,
|
| 1941 |
+
"step": 1380
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 40.73529411764706,
|
| 1945 |
+
"grad_norm": 0.13304731249809265,
|
| 1946 |
+
"learning_rate": 9.09351907982751e-06,
|
| 1947 |
+
"loss": 0.6569,
|
| 1948 |
+
"step": 1385
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 40.88235294117647,
|
| 1952 |
+
"grad_norm": 0.14108242094516754,
|
| 1953 |
+
"learning_rate": 8.968530505087582e-06,
|
| 1954 |
+
"loss": 0.6894,
|
| 1955 |
+
"step": 1390
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 41.029411764705884,
|
| 1959 |
+
"grad_norm": 0.13457804918289185,
|
| 1960 |
+
"learning_rate": 8.845295007929446e-06,
|
| 1961 |
+
"loss": 0.6814,
|
| 1962 |
+
"step": 1395
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 41.1764705882353,
|
| 1966 |
+
"grad_norm": 0.13951298594474792,
|
| 1967 |
+
"learning_rate": 8.7238242464349e-06,
|
| 1968 |
+
"loss": 0.6721,
|
| 1969 |
+
"step": 1400
|
| 1970 |
+
},
|
| 1971 |
+
{
|
| 1972 |
+
"epoch": 41.3235294117647,
|
| 1973 |
+
"grad_norm": 0.14416338503360748,
|
| 1974 |
+
"learning_rate": 8.604129711741706e-06,
|
| 1975 |
+
"loss": 0.6881,
|
| 1976 |
+
"step": 1405
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 41.470588235294116,
|
| 1980 |
+
"grad_norm": 0.13295041024684906,
|
| 1981 |
+
"learning_rate": 8.486222726956508e-06,
|
| 1982 |
+
"loss": 0.6624,
|
| 1983 |
+
"step": 1410
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 41.61764705882353,
|
| 1987 |
+
"grad_norm": 0.1342659443616867,
|
| 1988 |
+
"learning_rate": 8.370114446083686e-06,
|
| 1989 |
+
"loss": 0.6956,
|
| 1990 |
+
"step": 1415
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"epoch": 41.76470588235294,
|
| 1994 |
+
"grad_norm": 0.13162069022655487,
|
| 1995 |
+
"learning_rate": 8.255815852970153e-06,
|
| 1996 |
+
"loss": 0.6646,
|
| 1997 |
+
"step": 1420
|
| 1998 |
+
},
|
| 1999 |
+
{
|
| 2000 |
+
"epoch": 41.911764705882355,
|
| 2001 |
+
"grad_norm": 0.12931868433952332,
|
| 2002 |
+
"learning_rate": 8.143337760266331e-06,
|
| 2003 |
+
"loss": 0.6618,
|
| 2004 |
+
"step": 1425
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 42.05882352941177,
|
| 2008 |
+
"grad_norm": 0.13857227563858032,
|
| 2009 |
+
"learning_rate": 8.032690808403232e-06,
|
| 2010 |
+
"loss": 0.6672,
|
| 2011 |
+
"step": 1430
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"epoch": 42.205882352941174,
|
| 2015 |
+
"grad_norm": 0.13812746107578278,
|
| 2016 |
+
"learning_rate": 7.923885464585884e-06,
|
| 2017 |
+
"loss": 0.6866,
|
| 2018 |
+
"step": 1435
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 42.35294117647059,
|
| 2022 |
+
"grad_norm": 0.1503993570804596,
|
| 2023 |
+
"learning_rate": 7.816932021803154e-06,
|
| 2024 |
+
"loss": 0.6885,
|
| 2025 |
+
"step": 1440
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 42.5,
|
| 2029 |
+
"grad_norm": 0.13175919651985168,
|
| 2030 |
+
"learning_rate": 7.711840597853998e-06,
|
| 2031 |
+
"loss": 0.6686,
|
| 2032 |
+
"step": 1445
|
| 2033 |
+
},
|
| 2034 |
+
{
|
| 2035 |
+
"epoch": 42.64705882352941,
|
| 2036 |
+
"grad_norm": 0.13319700956344604,
|
| 2037 |
+
"learning_rate": 7.608621134390344e-06,
|
| 2038 |
+
"loss": 0.6561,
|
| 2039 |
+
"step": 1450
|
| 2040 |
+
},
|
| 2041 |
+
{
|
| 2042 |
+
"epoch": 42.794117647058826,
|
| 2043 |
+
"grad_norm": 0.1399184763431549,
|
| 2044 |
+
"learning_rate": 7.507283395976592e-06,
|
| 2045 |
+
"loss": 0.6537,
|
| 2046 |
+
"step": 1455
|
| 2047 |
+
},
|
| 2048 |
+
{
|
| 2049 |
+
"epoch": 42.94117647058823,
|
| 2050 |
+
"grad_norm": 0.13498006761074066,
|
| 2051 |
+
"learning_rate": 7.407836969165911e-06,
|
| 2052 |
+
"loss": 0.6886,
|
| 2053 |
+
"step": 1460
|
| 2054 |
+
},
|
| 2055 |
+
{
|
| 2056 |
+
"epoch": 43.088235294117645,
|
| 2057 |
+
"grad_norm": 0.1388946920633316,
|
| 2058 |
+
"learning_rate": 7.310291261593308e-06,
|
| 2059 |
+
"loss": 0.6797,
|
| 2060 |
+
"step": 1465
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"epoch": 43.23529411764706,
|
| 2064 |
+
"grad_norm": 0.13527587056159973,
|
| 2065 |
+
"learning_rate": 7.2146555010857155e-06,
|
| 2066 |
+
"loss": 0.6813,
|
| 2067 |
+
"step": 1470
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 43.38235294117647,
|
| 2071 |
+
"grad_norm": 0.13819634914398193,
|
| 2072 |
+
"learning_rate": 7.120938734789012e-06,
|
| 2073 |
+
"loss": 0.6752,
|
| 2074 |
+
"step": 1475
|
| 2075 |
+
},
|
| 2076 |
+
{
|
| 2077 |
+
"epoch": 43.529411764705884,
|
| 2078 |
+
"grad_norm": 0.12921284139156342,
|
| 2079 |
+
"learning_rate": 7.029149828312145e-06,
|
| 2080 |
+
"loss": 0.6761,
|
| 2081 |
+
"step": 1480
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 43.6764705882353,
|
| 2085 |
+
"grad_norm": 0.13316689431667328,
|
| 2086 |
+
"learning_rate": 6.93929746488849e-06,
|
| 2087 |
+
"loss": 0.6632,
|
| 2088 |
+
"step": 1485
|
| 2089 |
+
},
|
| 2090 |
+
{
|
| 2091 |
+
"epoch": 43.8235294117647,
|
| 2092 |
+
"grad_norm": 0.13092860579490662,
|
| 2093 |
+
"learning_rate": 6.851390144554372e-06,
|
| 2094 |
+
"loss": 0.6705,
|
| 2095 |
+
"step": 1490
|
| 2096 |
+
},
|
| 2097 |
+
{
|
| 2098 |
+
"epoch": 43.970588235294116,
|
| 2099 |
+
"grad_norm": 0.13236363232135773,
|
| 2100 |
+
"learning_rate": 6.765436183344996e-06,
|
| 2101 |
+
"loss": 0.6602,
|
| 2102 |
+
"step": 1495
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"epoch": 44.11764705882353,
|
| 2106 |
+
"grad_norm": 0.14111852645874023,
|
| 2107 |
+
"learning_rate": 6.6814437125077135e-06,
|
| 2108 |
+
"loss": 0.6554,
|
| 2109 |
+
"step": 1500
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 44.26470588235294,
|
| 2113 |
+
"grad_norm": 0.13777071237564087,
|
| 2114 |
+
"learning_rate": 6.599420677732848e-06,
|
| 2115 |
+
"loss": 0.6783,
|
| 2116 |
+
"step": 1505
|
| 2117 |
+
},
|
| 2118 |
+
{
|
| 2119 |
+
"epoch": 44.411764705882355,
|
| 2120 |
+
"grad_norm": 0.15148292481899261,
|
| 2121 |
+
"learning_rate": 6.519374838401997e-06,
|
| 2122 |
+
"loss": 0.6818,
|
| 2123 |
+
"step": 1510
|
| 2124 |
+
},
|
| 2125 |
+
{
|
| 2126 |
+
"epoch": 44.55882352941177,
|
| 2127 |
+
"grad_norm": 0.14915376901626587,
|
| 2128 |
+
"learning_rate": 6.44131376685401e-06,
|
| 2129 |
+
"loss": 0.6758,
|
| 2130 |
+
"step": 1515
|
| 2131 |
+
},
|
| 2132 |
+
{
|
| 2133 |
+
"epoch": 44.705882352941174,
|
| 2134 |
+
"grad_norm": 0.13784313201904297,
|
| 2135 |
+
"learning_rate": 6.36524484766865e-06,
|
| 2136 |
+
"loss": 0.6652,
|
| 2137 |
+
"step": 1520
|
| 2138 |
+
},
|
| 2139 |
+
{
|
| 2140 |
+
"epoch": 44.85294117647059,
|
| 2141 |
+
"grad_norm": 0.13148203492164612,
|
| 2142 |
+
"learning_rate": 6.291175276968002e-06,
|
| 2143 |
+
"loss": 0.6758,
|
| 2144 |
+
"step": 1525
|
| 2145 |
+
},
|
| 2146 |
+
{
|
| 2147 |
+
"epoch": 45.0,
|
| 2148 |
+
"grad_norm": 0.14285942912101746,
|
| 2149 |
+
"learning_rate": 6.219112061735721e-06,
|
| 2150 |
+
"loss": 0.6716,
|
| 2151 |
+
"step": 1530
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 45.14705882352941,
|
| 2155 |
+
"grad_norm": 0.14051543176174164,
|
| 2156 |
+
"learning_rate": 6.149062019154174e-06,
|
| 2157 |
+
"loss": 0.6833,
|
| 2158 |
+
"step": 1535
|
| 2159 |
+
},
|
| 2160 |
+
{
|
| 2161 |
+
"epoch": 45.294117647058826,
|
| 2162 |
+
"grad_norm": 0.12927637994289398,
|
| 2163 |
+
"learning_rate": 6.081031775959542e-06,
|
| 2164 |
+
"loss": 0.648,
|
| 2165 |
+
"step": 1540
|
| 2166 |
+
},
|
| 2167 |
+
{
|
| 2168 |
+
"epoch": 45.44117647058823,
|
| 2169 |
+
"grad_norm": 0.1433332860469818,
|
| 2170 |
+
"learning_rate": 6.0150277678149055e-06,
|
| 2171 |
+
"loss": 0.6377,
|
| 2172 |
+
"step": 1545
|
| 2173 |
+
},
|
| 2174 |
+
{
|
| 2175 |
+
"epoch": 45.588235294117645,
|
| 2176 |
+
"grad_norm": 0.1298450231552124,
|
| 2177 |
+
"learning_rate": 5.951056238701456e-06,
|
| 2178 |
+
"loss": 0.6866,
|
| 2179 |
+
"step": 1550
|
| 2180 |
+
},
|
| 2181 |
+
{
|
| 2182 |
+
"epoch": 45.73529411764706,
|
| 2183 |
+
"grad_norm": 0.13296933472156525,
|
| 2184 |
+
"learning_rate": 5.889123240327819e-06,
|
| 2185 |
+
"loss": 0.6747,
|
| 2186 |
+
"step": 1555
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 45.88235294117647,
|
| 2190 |
+
"grad_norm": 0.1386398822069168,
|
| 2191 |
+
"learning_rate": 5.829234631557524e-06,
|
| 2192 |
+
"loss": 0.6827,
|
| 2193 |
+
"step": 1560
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 46.029411764705884,
|
| 2197 |
+
"grad_norm": 0.12945467233657837,
|
| 2198 |
+
"learning_rate": 5.771396077854802e-06,
|
| 2199 |
+
"loss": 0.6823,
|
| 2200 |
+
"step": 1565
|
| 2201 |
+
},
|
| 2202 |
+
{
|
| 2203 |
+
"epoch": 46.1764705882353,
|
| 2204 |
+
"grad_norm": 0.1447058618068695,
|
| 2205 |
+
"learning_rate": 5.715613050748604e-06,
|
| 2206 |
+
"loss": 0.6542,
|
| 2207 |
+
"step": 1570
|
| 2208 |
+
},
|
| 2209 |
+
{
|
| 2210 |
+
"epoch": 46.3235294117647,
|
| 2211 |
+
"grad_norm": 0.13116587698459625,
|
| 2212 |
+
"learning_rate": 5.661890827315004e-06,
|
| 2213 |
+
"loss": 0.664,
|
| 2214 |
+
"step": 1575
|
| 2215 |
+
},
|
| 2216 |
+
{
|
| 2217 |
+
"epoch": 46.470588235294116,
|
| 2218 |
+
"grad_norm": 0.13510237634181976,
|
| 2219 |
+
"learning_rate": 5.61023448967798e-06,
|
| 2220 |
+
"loss": 0.6698,
|
| 2221 |
+
"step": 1580
|
| 2222 |
+
},
|
| 2223 |
+
{
|
| 2224 |
+
"epoch": 46.61764705882353,
|
| 2225 |
+
"grad_norm": 0.13816939294338226,
|
| 2226 |
+
"learning_rate": 5.560648924528657e-06,
|
| 2227 |
+
"loss": 0.7097,
|
| 2228 |
+
"step": 1585
|
| 2229 |
+
},
|
| 2230 |
+
{
|
| 2231 |
+
"epoch": 46.76470588235294,
|
| 2232 |
+
"grad_norm": 0.1419830620288849,
|
| 2233 |
+
"learning_rate": 5.513138822663016e-06,
|
| 2234 |
+
"loss": 0.6905,
|
| 2235 |
+
"step": 1590
|
| 2236 |
+
},
|
| 2237 |
+
{
|
| 2238 |
+
"epoch": 46.911764705882355,
|
| 2239 |
+
"grad_norm": 0.13188646733760834,
|
| 2240 |
+
"learning_rate": 5.467708678538148e-06,
|
| 2241 |
+
"loss": 0.6457,
|
| 2242 |
+
"step": 1595
|
| 2243 |
+
},
|
| 2244 |
+
{
|
| 2245 |
+
"epoch": 47.05882352941177,
|
| 2246 |
+
"grad_norm": 0.1413869559764862,
|
| 2247 |
+
"learning_rate": 5.424362789847082e-06,
|
| 2248 |
+
"loss": 0.6766,
|
| 2249 |
+
"step": 1600
|
| 2250 |
+
},
|
| 2251 |
+
{
|
| 2252 |
+
"epoch": 47.205882352941174,
|
| 2253 |
+
"grad_norm": 0.13414201140403748,
|
| 2254 |
+
"learning_rate": 5.38310525711221e-06,
|
| 2255 |
+
"loss": 0.6795,
|
| 2256 |
+
"step": 1605
|
| 2257 |
+
},
|
| 2258 |
+
{
|
| 2259 |
+
"epoch": 47.35294117647059,
|
| 2260 |
+
"grad_norm": 0.1363234668970108,
|
| 2261 |
+
"learning_rate": 5.343939983297398e-06,
|
| 2262 |
+
"loss": 0.6713,
|
| 2263 |
+
"step": 1610
|
| 2264 |
+
},
|
| 2265 |
+
{
|
| 2266 |
+
"epoch": 47.5,
|
| 2267 |
+
"grad_norm": 0.1344790905714035,
|
| 2268 |
+
"learning_rate": 5.3068706734387484e-06,
|
| 2269 |
+
"loss": 0.6584,
|
| 2270 |
+
"step": 1615
|
| 2271 |
+
},
|
| 2272 |
+
{
|
| 2273 |
+
"epoch": 47.64705882352941,
|
| 2274 |
+
"grad_norm": 0.1452033668756485,
|
| 2275 |
+
"learning_rate": 5.271900834294105e-06,
|
| 2276 |
+
"loss": 0.667,
|
| 2277 |
+
"step": 1620
|
| 2278 |
+
},
|
| 2279 |
+
{
|
| 2280 |
+
"epoch": 47.794117647058826,
|
| 2281 |
+
"grad_norm": 0.13405530154705048,
|
| 2282 |
+
"learning_rate": 5.239033774011322e-06,
|
| 2283 |
+
"loss": 0.669,
|
| 2284 |
+
"step": 1625
|
| 2285 |
+
},
|
| 2286 |
+
{
|
| 2287 |
+
"epoch": 47.94117647058823,
|
| 2288 |
+
"grad_norm": 0.13740584254264832,
|
| 2289 |
+
"learning_rate": 5.208272601815313e-06,
|
| 2290 |
+
"loss": 0.6836,
|
| 2291 |
+
"step": 1630
|
| 2292 |
+
},
|
| 2293 |
+
{
|
| 2294 |
+
"epoch": 48.088235294117645,
|
| 2295 |
+
"grad_norm": 0.12757079303264618,
|
| 2296 |
+
"learning_rate": 5.1796202277139075e-06,
|
| 2297 |
+
"loss": 0.6909,
|
| 2298 |
+
"step": 1635
|
| 2299 |
+
},
|
| 2300 |
+
{
|
| 2301 |
+
"epoch": 48.23529411764706,
|
| 2302 |
+
"grad_norm": 0.12381689995527267,
|
| 2303 |
+
"learning_rate": 5.1530793622225725e-06,
|
| 2304 |
+
"loss": 0.6605,
|
| 2305 |
+
"step": 1640
|
| 2306 |
+
},
|
| 2307 |
+
{
|
| 2308 |
+
"epoch": 48.38235294117647,
|
| 2309 |
+
"grad_norm": 0.13950037956237793,
|
| 2310 |
+
"learning_rate": 5.128652516107996e-06,
|
| 2311 |
+
"loss": 0.6814,
|
| 2312 |
+
"step": 1645
|
| 2313 |
+
},
|
| 2314 |
+
{
|
| 2315 |
+
"epoch": 48.529411764705884,
|
| 2316 |
+
"grad_norm": 0.13073165714740753,
|
| 2317 |
+
"learning_rate": 5.10634200015057e-06,
|
| 2318 |
+
"loss": 0.6866,
|
| 2319 |
+
"step": 1650
|
| 2320 |
+
},
|
| 2321 |
+
{
|
| 2322 |
+
"epoch": 48.6764705882353,
|
| 2323 |
+
"grad_norm": 0.1424126774072647,
|
| 2324 |
+
"learning_rate": 5.086149924925788e-06,
|
| 2325 |
+
"loss": 0.6697,
|
| 2326 |
+
"step": 1655
|
| 2327 |
+
},
|
| 2328 |
+
{
|
| 2329 |
+
"epoch": 48.8235294117647,
|
| 2330 |
+
"grad_norm": 0.15078318119049072,
|
| 2331 |
+
"learning_rate": 5.068078200604584e-06,
|
| 2332 |
+
"loss": 0.6615,
|
| 2333 |
+
"step": 1660
|
| 2334 |
+
},
|
| 2335 |
+
{
|
| 2336 |
+
"epoch": 48.970588235294116,
|
| 2337 |
+
"grad_norm": 0.1373935043811798,
|
| 2338 |
+
"learning_rate": 5.052128536772629e-06,
|
| 2339 |
+
"loss": 0.6665,
|
| 2340 |
+
"step": 1665
|
| 2341 |
+
},
|
| 2342 |
+
{
|
| 2343 |
+
"epoch": 49.11764705882353,
|
| 2344 |
+
"grad_norm": 0.1401982456445694,
|
| 2345 |
+
"learning_rate": 5.038302442268617e-06,
|
| 2346 |
+
"loss": 0.6597,
|
| 2347 |
+
"step": 1670
|
| 2348 |
+
},
|
| 2349 |
+
{
|
| 2350 |
+
"epoch": 49.26470588235294,
|
| 2351 |
+
"grad_norm": 0.1401190608739853,
|
| 2352 |
+
"learning_rate": 5.026601225041503e-06,
|
| 2353 |
+
"loss": 0.6929,
|
| 2354 |
+
"step": 1675
|
| 2355 |
+
},
|
| 2356 |
+
{
|
| 2357 |
+
"epoch": 49.411764705882355,
|
| 2358 |
+
"grad_norm": 0.1368054449558258,
|
| 2359 |
+
"learning_rate": 5.0170259920268025e-06,
|
| 2360 |
+
"loss": 0.6923,
|
| 2361 |
+
"step": 1680
|
| 2362 |
+
},
|
| 2363 |
+
{
|
| 2364 |
+
"epoch": 49.55882352941177,
|
| 2365 |
+
"grad_norm": 0.13621118664741516,
|
| 2366 |
+
"learning_rate": 5.009577649041847e-06,
|
| 2367 |
+
"loss": 0.6574,
|
| 2368 |
+
"step": 1685
|
| 2369 |
+
},
|
| 2370 |
+
{
|
| 2371 |
+
"epoch": 49.705882352941174,
|
| 2372 |
+
"grad_norm": 0.13294735550880432,
|
| 2373 |
+
"learning_rate": 5.004256900700115e-06,
|
| 2374 |
+
"loss": 0.6646,
|
| 2375 |
+
"step": 1690
|
| 2376 |
+
},
|
| 2377 |
+
{
|
| 2378 |
+
"epoch": 49.85294117647059,
|
| 2379 |
+
"grad_norm": 0.14144855737686157,
|
| 2380 |
+
"learning_rate": 5.001064250344557e-06,
|
| 2381 |
+
"loss": 0.666,
|
| 2382 |
+
"step": 1695
|
| 2383 |
+
},
|
| 2384 |
+
{
|
| 2385 |
+
"epoch": 50.0,
|
| 2386 |
+
"grad_norm": 0.13993091881275177,
|
| 2387 |
+
"learning_rate": 5e-06,
|
| 2388 |
+
"loss": 0.6593,
|
| 2389 |
+
"step": 1700
|
| 2390 |
+
},
|
| 2391 |
+
{
|
| 2392 |
+
"epoch": 50.0,
|
| 2393 |
+
"step": 1700,
|
| 2394 |
+
"total_flos": 2.629944131882844e+18,
|
| 2395 |
+
"train_loss": 0.6986723100437837,
|
| 2396 |
+
"train_runtime": 41334.8641,
|
| 2397 |
+
"train_samples_per_second": 0.481,
|
| 2398 |
+
"train_steps_per_second": 0.041
|
| 2399 |
+
}
|
| 2400 |
+
],
|
| 2401 |
+
"logging_steps": 5,
|
| 2402 |
+
"max_steps": 1700,
|
| 2403 |
+
"num_input_tokens_seen": 0,
|
| 2404 |
+
"num_train_epochs": 50,
|
| 2405 |
+
"save_steps": 500,
|
| 2406 |
+
"stateful_callbacks": {
|
| 2407 |
+
"TrainerControl": {
|
| 2408 |
+
"args": {
|
| 2409 |
+
"should_epoch_stop": false,
|
| 2410 |
+
"should_evaluate": false,
|
| 2411 |
+
"should_log": false,
|
| 2412 |
+
"should_save": true,
|
| 2413 |
+
"should_training_stop": true
|
| 2414 |
+
},
|
| 2415 |
+
"attributes": {}
|
| 2416 |
+
}
|
| 2417 |
+
},
|
| 2418 |
+
"total_flos": 2.629944131882844e+18,
|
| 2419 |
+
"train_batch_size": 6,
|
| 2420 |
+
"trial_name": null,
|
| 2421 |
+
"trial_params": null
|
| 2422 |
+
}
|