Model save
Browse files- README.md +67 -0
- all_results.json +9 -0
- train_results.json +9 -0
- trainer_state.json +1232 -0
README.md
ADDED
|
@@ -0,0 +1,67 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
base_model: google/gemma-7b
|
| 3 |
+
library_name: transformers
|
| 4 |
+
model_name: gemma-7b-silvia-shuffled-5e-5-05-v6-1sftt
|
| 5 |
+
tags:
|
| 6 |
+
- generated_from_trainer
|
| 7 |
+
- trl
|
| 8 |
+
- orpo
|
| 9 |
+
licence: license
|
| 10 |
+
---
|
| 11 |
+
|
| 12 |
+
# Model Card for gemma-7b-silvia-shuffled-5e-5-05-v6-1sftt
|
| 13 |
+
|
| 14 |
+
This model is a fine-tuned version of [google/gemma-7b](https://huggingface.co/google/gemma-7b).
|
| 15 |
+
It has been trained using [TRL](https://github.com/huggingface/trl).
|
| 16 |
+
|
| 17 |
+
## Quick start
|
| 18 |
+
|
| 19 |
+
```python
|
| 20 |
+
from transformers import pipeline
|
| 21 |
+
|
| 22 |
+
question = "If you had a time machine, but could only go to the past or the future once and never return, which would you choose and why?"
|
| 23 |
+
generator = pipeline("text-generation", model="silviasapora/gemma-7b-silvia-shuffled-5e-5-05-v6-1sftt", 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/silvias/huggingface/runs/et5ungxj)
|
| 31 |
+
|
| 32 |
+
|
| 33 |
+
This model was trained with ORPO, a method introduced in [ORPO: Monolithic Preference Optimization without Reference Model](https://huggingface.co/papers/2403.07691).
|
| 34 |
+
|
| 35 |
+
### Framework versions
|
| 36 |
+
|
| 37 |
+
- TRL: 0.13.0
|
| 38 |
+
- Transformers: 4.48.1
|
| 39 |
+
- Pytorch: 2.5.1
|
| 40 |
+
- Datasets: 3.2.0
|
| 41 |
+
- Tokenizers: 0.21.0
|
| 42 |
+
|
| 43 |
+
## Citations
|
| 44 |
+
|
| 45 |
+
Cite ORPO as:
|
| 46 |
+
|
| 47 |
+
```bibtex
|
| 48 |
+
@article{hong2024orpo,
|
| 49 |
+
title = {{ORPO: Monolithic Preference Optimization without Reference Model}},
|
| 50 |
+
author = {Jiwoo Hong and Noah Lee and James Thorne},
|
| 51 |
+
year = 2024,
|
| 52 |
+
eprint = {arXiv:2403.07691}
|
| 53 |
+
}
|
| 54 |
+
```
|
| 55 |
+
|
| 56 |
+
Cite TRL as:
|
| 57 |
+
|
| 58 |
+
```bibtex
|
| 59 |
+
@misc{vonwerra2022trl,
|
| 60 |
+
title = {{TRL: Transformer Reinforcement Learning}},
|
| 61 |
+
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},
|
| 62 |
+
year = 2020,
|
| 63 |
+
journal = {GitHub repository},
|
| 64 |
+
publisher = {GitHub},
|
| 65 |
+
howpublished = {\url{https://github.com/huggingface/trl}}
|
| 66 |
+
}
|
| 67 |
+
```
|
all_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 2.981333333333333,
|
| 3 |
+
"total_flos": 0.0,
|
| 4 |
+
"train_loss": 54.80183921716152,
|
| 5 |
+
"train_runtime": 10133.5503,
|
| 6 |
+
"train_samples": 7500,
|
| 7 |
+
"train_samples_per_second": 2.22,
|
| 8 |
+
"train_steps_per_second": 0.035
|
| 9 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 2.981333333333333,
|
| 3 |
+
"total_flos": 0.0,
|
| 4 |
+
"train_loss": 54.80183921716152,
|
| 5 |
+
"train_runtime": 10133.5503,
|
| 6 |
+
"train_samples": 7500,
|
| 7 |
+
"train_samples_per_second": 2.22,
|
| 8 |
+
"train_steps_per_second": 0.035
|
| 9 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1232 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 2.981333333333333,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 351,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.042666666666666665,
|
| 13 |
+
"grad_norm": 1961.1512451171875,
|
| 14 |
+
"learning_rate": 6.944444444444445e-06,
|
| 15 |
+
"log_odds_chosen": 3.233767032623291,
|
| 16 |
+
"log_odds_ratio": -7.3191938400268555,
|
| 17 |
+
"logps/chosen": -23.426687240600586,
|
| 18 |
+
"logps/rejected": -26.660680770874023,
|
| 19 |
+
"loss": 425.3708,
|
| 20 |
+
"nll_loss": 9.728386878967285,
|
| 21 |
+
"po_loss": 3.5644500255584717,
|
| 22 |
+
"rewards/accuracies": 0.515625,
|
| 23 |
+
"rewards/chosen": -11.713343620300293,
|
| 24 |
+
"rewards/margins": 1.6169971227645874,
|
| 25 |
+
"rewards/rejected": -13.330340385437012,
|
| 26 |
+
"step": 5
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.08533333333333333,
|
| 30 |
+
"grad_norm": 500.5514831542969,
|
| 31 |
+
"learning_rate": 1.388888888888889e-05,
|
| 32 |
+
"log_odds_chosen": 3.315504550933838,
|
| 33 |
+
"log_odds_ratio": -6.866879463195801,
|
| 34 |
+
"logps/chosen": -23.438692092895508,
|
| 35 |
+
"logps/rejected": -26.755935668945312,
|
| 36 |
+
"loss": 393.0612,
|
| 37 |
+
"nll_loss": 8.955938339233398,
|
| 38 |
+
"po_loss": 3.3272223472595215,
|
| 39 |
+
"rewards/accuracies": 0.48124998807907104,
|
| 40 |
+
"rewards/chosen": -11.719346046447754,
|
| 41 |
+
"rewards/margins": 1.658623456954956,
|
| 42 |
+
"rewards/rejected": -13.377967834472656,
|
| 43 |
+
"step": 10
|
| 44 |
+
},
|
| 45 |
+
{
|
| 46 |
+
"epoch": 0.128,
|
| 47 |
+
"grad_norm": 514.0573120117188,
|
| 48 |
+
"learning_rate": 2.0833333333333336e-05,
|
| 49 |
+
"log_odds_chosen": 7.634564399719238,
|
| 50 |
+
"log_odds_ratio": -6.296060562133789,
|
| 51 |
+
"logps/chosen": -22.5439395904541,
|
| 52 |
+
"logps/rejected": -30.179019927978516,
|
| 53 |
+
"loss": 378.1973,
|
| 54 |
+
"nll_loss": 8.766839981079102,
|
| 55 |
+
"po_loss": 3.0518250465393066,
|
| 56 |
+
"rewards/accuracies": 0.53125,
|
| 57 |
+
"rewards/chosen": -11.27196979522705,
|
| 58 |
+
"rewards/margins": 3.8175387382507324,
|
| 59 |
+
"rewards/rejected": -15.089509963989258,
|
| 60 |
+
"step": 15
|
| 61 |
+
},
|
| 62 |
+
{
|
| 63 |
+
"epoch": 0.17066666666666666,
|
| 64 |
+
"grad_norm": 3820.703125,
|
| 65 |
+
"learning_rate": 2.777777777777778e-05,
|
| 66 |
+
"log_odds_chosen": -1.3400392532348633,
|
| 67 |
+
"log_odds_ratio": -8.175825119018555,
|
| 68 |
+
"logps/chosen": -18.689817428588867,
|
| 69 |
+
"logps/rejected": -17.35008430480957,
|
| 70 |
+
"loss": 363.51,
|
| 71 |
+
"nll_loss": 7.390625953674316,
|
| 72 |
+
"po_loss": 3.9690635204315186,
|
| 73 |
+
"rewards/accuracies": 0.484375,
|
| 74 |
+
"rewards/chosen": -9.344908714294434,
|
| 75 |
+
"rewards/margins": -0.6698670983314514,
|
| 76 |
+
"rewards/rejected": -8.675042152404785,
|
| 77 |
+
"step": 20
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 0.21333333333333335,
|
| 81 |
+
"grad_norm": 417.3852844238281,
|
| 82 |
+
"learning_rate": 3.472222222222222e-05,
|
| 83 |
+
"log_odds_chosen": -0.37303870916366577,
|
| 84 |
+
"log_odds_ratio": -2.343538284301758,
|
| 85 |
+
"logps/chosen": -5.883327484130859,
|
| 86 |
+
"logps/rejected": -5.492794036865234,
|
| 87 |
+
"loss": 134.8965,
|
| 88 |
+
"nll_loss": 3.241476535797119,
|
| 89 |
+
"po_loss": 0.9740395545959473,
|
| 90 |
+
"rewards/accuracies": 0.5562499761581421,
|
| 91 |
+
"rewards/chosen": -2.9416637420654297,
|
| 92 |
+
"rewards/margins": -0.19526681303977966,
|
| 93 |
+
"rewards/rejected": -2.746397018432617,
|
| 94 |
+
"step": 25
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.256,
|
| 98 |
+
"grad_norm": 330.9832458496094,
|
| 99 |
+
"learning_rate": 4.166666666666667e-05,
|
| 100 |
+
"log_odds_chosen": 0.3532889187335968,
|
| 101 |
+
"log_odds_ratio": -0.7218837141990662,
|
| 102 |
+
"logps/chosen": -1.7468492984771729,
|
| 103 |
+
"logps/rejected": -2.0471103191375732,
|
| 104 |
+
"loss": 70.938,
|
| 105 |
+
"nll_loss": 2.1013693809509277,
|
| 106 |
+
"po_loss": 0.11544491350650787,
|
| 107 |
+
"rewards/accuracies": 0.625,
|
| 108 |
+
"rewards/chosen": -0.8734246492385864,
|
| 109 |
+
"rewards/margins": 0.15013054013252258,
|
| 110 |
+
"rewards/rejected": -1.0235551595687866,
|
| 111 |
+
"step": 30
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 0.2986666666666667,
|
| 115 |
+
"grad_norm": 284.9355163574219,
|
| 116 |
+
"learning_rate": 4.8611111111111115e-05,
|
| 117 |
+
"log_odds_chosen": 0.4371984004974365,
|
| 118 |
+
"log_odds_ratio": -0.6356752514839172,
|
| 119 |
+
"logps/chosen": -1.627811074256897,
|
| 120 |
+
"logps/rejected": -1.993390679359436,
|
| 121 |
+
"loss": 63.8055,
|
| 122 |
+
"nll_loss": 1.9257453680038452,
|
| 123 |
+
"po_loss": 0.06817711889743805,
|
| 124 |
+
"rewards/accuracies": 0.6937500238418579,
|
| 125 |
+
"rewards/chosen": -0.8139055371284485,
|
| 126 |
+
"rewards/margins": 0.18278983235359192,
|
| 127 |
+
"rewards/rejected": -0.996695339679718,
|
| 128 |
+
"step": 35
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.3413333333333333,
|
| 132 |
+
"grad_norm": 126.80204772949219,
|
| 133 |
+
"learning_rate": 4.998010925565448e-05,
|
| 134 |
+
"log_odds_chosen": 0.36739838123321533,
|
| 135 |
+
"log_odds_ratio": -0.6945838928222656,
|
| 136 |
+
"logps/chosen": -1.696966528892517,
|
| 137 |
+
"logps/rejected": -2.019742488861084,
|
| 138 |
+
"loss": 66.0958,
|
| 139 |
+
"nll_loss": 1.972928762435913,
|
| 140 |
+
"po_loss": 0.0925649031996727,
|
| 141 |
+
"rewards/accuracies": 0.637499988079071,
|
| 142 |
+
"rewards/chosen": -0.8484832644462585,
|
| 143 |
+
"rewards/margins": 0.16138817369937897,
|
| 144 |
+
"rewards/rejected": -1.009871244430542,
|
| 145 |
+
"step": 40
|
| 146 |
+
},
|
| 147 |
+
{
|
| 148 |
+
"epoch": 0.384,
|
| 149 |
+
"grad_norm": 734.975830078125,
|
| 150 |
+
"learning_rate": 4.989935734988098e-05,
|
| 151 |
+
"log_odds_chosen": 0.2258051633834839,
|
| 152 |
+
"log_odds_ratio": -0.6970190405845642,
|
| 153 |
+
"logps/chosen": -1.5770127773284912,
|
| 154 |
+
"logps/rejected": -1.7601944208145142,
|
| 155 |
+
"loss": 61.9986,
|
| 156 |
+
"nll_loss": 1.8443195819854736,
|
| 157 |
+
"po_loss": 0.09313799440860748,
|
| 158 |
+
"rewards/accuracies": 0.6000000238418579,
|
| 159 |
+
"rewards/chosen": -0.7885063886642456,
|
| 160 |
+
"rewards/margins": 0.09159073978662491,
|
| 161 |
+
"rewards/rejected": -0.8800972104072571,
|
| 162 |
+
"step": 45
|
| 163 |
+
},
|
| 164 |
+
{
|
| 165 |
+
"epoch": 0.4266666666666667,
|
| 166 |
+
"grad_norm": 98.6443862915039,
|
| 167 |
+
"learning_rate": 4.975670171853926e-05,
|
| 168 |
+
"log_odds_chosen": 0.3061736226081848,
|
| 169 |
+
"log_odds_ratio": -0.6558084487915039,
|
| 170 |
+
"logps/chosen": -1.410742998123169,
|
| 171 |
+
"logps/rejected": -1.6612850427627563,
|
| 172 |
+
"loss": 56.1342,
|
| 173 |
+
"nll_loss": 1.6897376775741577,
|
| 174 |
+
"po_loss": 0.06445512920618057,
|
| 175 |
+
"rewards/accuracies": 0.6343749761581421,
|
| 176 |
+
"rewards/chosen": -0.7053714990615845,
|
| 177 |
+
"rewards/margins": 0.1252710521221161,
|
| 178 |
+
"rewards/rejected": -0.8306425213813782,
|
| 179 |
+
"step": 50
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 0.4693333333333333,
|
| 183 |
+
"grad_norm": 138.6815948486328,
|
| 184 |
+
"learning_rate": 4.9552497026005974e-05,
|
| 185 |
+
"log_odds_chosen": 0.21871152520179749,
|
| 186 |
+
"log_odds_ratio": -0.7383303046226501,
|
| 187 |
+
"logps/chosen": -1.4634255170822144,
|
| 188 |
+
"logps/rejected": -1.6147739887237549,
|
| 189 |
+
"loss": 59.4312,
|
| 190 |
+
"nll_loss": 1.7589048147201538,
|
| 191 |
+
"po_loss": 0.09831906855106354,
|
| 192 |
+
"rewards/accuracies": 0.581250011920929,
|
| 193 |
+
"rewards/chosen": -0.7317127585411072,
|
| 194 |
+
"rewards/margins": 0.07567427307367325,
|
| 195 |
+
"rewards/rejected": -0.8073869943618774,
|
| 196 |
+
"step": 55
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"epoch": 0.512,
|
| 200 |
+
"grad_norm": 102.89739227294922,
|
| 201 |
+
"learning_rate": 4.928725095732169e-05,
|
| 202 |
+
"log_odds_chosen": 0.401327520608902,
|
| 203 |
+
"log_odds_ratio": -0.6062948107719421,
|
| 204 |
+
"logps/chosen": -1.2786072492599487,
|
| 205 |
+
"logps/rejected": -1.585089921951294,
|
| 206 |
+
"loss": 51.5199,
|
| 207 |
+
"nll_loss": 1.5874063968658447,
|
| 208 |
+
"po_loss": 0.02259092405438423,
|
| 209 |
+
"rewards/accuracies": 0.6812499761581421,
|
| 210 |
+
"rewards/chosen": -0.6393036246299744,
|
| 211 |
+
"rewards/margins": 0.15324126183986664,
|
| 212 |
+
"rewards/rejected": -0.792544960975647,
|
| 213 |
+
"step": 60
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.5546666666666666,
|
| 217 |
+
"grad_norm": 100.74688720703125,
|
| 218 |
+
"learning_rate": 4.896162295600589e-05,
|
| 219 |
+
"log_odds_chosen": 0.33866989612579346,
|
| 220 |
+
"log_odds_ratio": -0.6249087452888489,
|
| 221 |
+
"logps/chosen": -1.3379658460617065,
|
| 222 |
+
"logps/rejected": -1.6066213846206665,
|
| 223 |
+
"loss": 53.1437,
|
| 224 |
+
"nll_loss": 1.6211912631988525,
|
| 225 |
+
"po_loss": 0.03954792767763138,
|
| 226 |
+
"rewards/accuracies": 0.675000011920929,
|
| 227 |
+
"rewards/chosen": -0.6689829230308533,
|
| 228 |
+
"rewards/margins": 0.13432787358760834,
|
| 229 |
+
"rewards/rejected": -0.8033106923103333,
|
| 230 |
+
"step": 65
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 0.5973333333333334,
|
| 234 |
+
"grad_norm": 107.1102523803711,
|
| 235 |
+
"learning_rate": 4.8576422584576514e-05,
|
| 236 |
+
"log_odds_chosen": 0.28569236397743225,
|
| 237 |
+
"log_odds_ratio": -0.6388062238693237,
|
| 238 |
+
"logps/chosen": -1.2627140283584595,
|
| 239 |
+
"logps/rejected": -1.478611707687378,
|
| 240 |
+
"loss": 51.3954,
|
| 241 |
+
"nll_loss": 1.5717823505401611,
|
| 242 |
+
"po_loss": 0.034325726330280304,
|
| 243 |
+
"rewards/accuracies": 0.628125011920929,
|
| 244 |
+
"rewards/chosen": -0.6313570141792297,
|
| 245 |
+
"rewards/margins": 0.10794880241155624,
|
| 246 |
+
"rewards/rejected": -0.739305853843689,
|
| 247 |
+
"step": 70
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.64,
|
| 251 |
+
"grad_norm": 87.37519073486328,
|
| 252 |
+
"learning_rate": 4.813260751184992e-05,
|
| 253 |
+
"log_odds_chosen": 0.4192368984222412,
|
| 254 |
+
"log_odds_ratio": -0.6039353609085083,
|
| 255 |
+
"logps/chosen": -1.2572083473205566,
|
| 256 |
+
"logps/rejected": -1.5851625204086304,
|
| 257 |
+
"loss": 50.7092,
|
| 258 |
+
"nll_loss": 1.5785938501358032,
|
| 259 |
+
"po_loss": 0.00606974633410573,
|
| 260 |
+
"rewards/accuracies": 0.6499999761581421,
|
| 261 |
+
"rewards/chosen": -0.6286041736602783,
|
| 262 |
+
"rewards/margins": 0.1639770269393921,
|
| 263 |
+
"rewards/rejected": -0.7925812602043152,
|
| 264 |
+
"step": 75
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 0.6826666666666666,
|
| 268 |
+
"grad_norm": 84.41800689697266,
|
| 269 |
+
"learning_rate": 4.763128113202537e-05,
|
| 270 |
+
"log_odds_chosen": 0.3978249132633209,
|
| 271 |
+
"log_odds_ratio": -0.6257617473602295,
|
| 272 |
+
"logps/chosen": -1.1973122358322144,
|
| 273 |
+
"logps/rejected": -1.5119242668151855,
|
| 274 |
+
"loss": 49.6894,
|
| 275 |
+
"nll_loss": 1.5280753374099731,
|
| 276 |
+
"po_loss": 0.024718867614865303,
|
| 277 |
+
"rewards/accuracies": 0.6031249761581421,
|
| 278 |
+
"rewards/chosen": -0.5986561179161072,
|
| 279 |
+
"rewards/margins": 0.1573060303926468,
|
| 280 |
+
"rewards/rejected": -0.7559621334075928,
|
| 281 |
+
"step": 80
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 0.7253333333333334,
|
| 285 |
+
"grad_norm": 217.4127655029297,
|
| 286 |
+
"learning_rate": 4.707368982147318e-05,
|
| 287 |
+
"log_odds_chosen": 0.9764216542243958,
|
| 288 |
+
"log_odds_ratio": -0.5953387022018433,
|
| 289 |
+
"logps/chosen": -1.2593096494674683,
|
| 290 |
+
"logps/rejected": -2.1358559131622314,
|
| 291 |
+
"loss": 50.7417,
|
| 292 |
+
"nll_loss": 1.5718034505844116,
|
| 293 |
+
"po_loss": 0.013875389471650124,
|
| 294 |
+
"rewards/accuracies": 0.628125011920929,
|
| 295 |
+
"rewards/chosen": -0.6296548247337341,
|
| 296 |
+
"rewards/margins": 0.43827304244041443,
|
| 297 |
+
"rewards/rejected": -1.0679279565811157,
|
| 298 |
+
"step": 85
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"epoch": 0.768,
|
| 302 |
+
"grad_norm": 74.66828918457031,
|
| 303 |
+
"learning_rate": 4.6461219840046654e-05,
|
| 304 |
+
"log_odds_chosen": 0.7273199558258057,
|
| 305 |
+
"log_odds_ratio": -0.6018415093421936,
|
| 306 |
+
"logps/chosen": -1.213512897491455,
|
| 307 |
+
"logps/rejected": -1.8320105075836182,
|
| 308 |
+
"loss": 48.283,
|
| 309 |
+
"nll_loss": 1.5248548984527588,
|
| 310 |
+
"po_loss": -0.016011018306016922,
|
| 311 |
+
"rewards/accuracies": 0.6625000238418579,
|
| 312 |
+
"rewards/chosen": -0.6067564487457275,
|
| 313 |
+
"rewards/margins": 0.30924877524375916,
|
| 314 |
+
"rewards/rejected": -0.9160052537918091,
|
| 315 |
+
"step": 90
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"epoch": 0.8106666666666666,
|
| 319 |
+
"grad_norm": 96.17576599121094,
|
| 320 |
+
"learning_rate": 4.579539388462173e-05,
|
| 321 |
+
"log_odds_chosen": 0.968981921672821,
|
| 322 |
+
"log_odds_ratio": -0.5852179527282715,
|
| 323 |
+
"logps/chosen": -1.1610548496246338,
|
| 324 |
+
"logps/rejected": -2.0138120651245117,
|
| 325 |
+
"loss": 46.2603,
|
| 326 |
+
"nll_loss": 1.5272886753082275,
|
| 327 |
+
"po_loss": -0.08165426552295685,
|
| 328 |
+
"rewards/accuracies": 0.637499988079071,
|
| 329 |
+
"rewards/chosen": -0.5805274248123169,
|
| 330 |
+
"rewards/margins": 0.4263785481452942,
|
| 331 |
+
"rewards/rejected": -1.0069060325622559,
|
| 332 |
+
"step": 95
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.8533333333333334,
|
| 336 |
+
"grad_norm": 1037.95703125,
|
| 337 |
+
"learning_rate": 4.5077867303432546e-05,
|
| 338 |
+
"log_odds_chosen": 2.305642604827881,
|
| 339 |
+
"log_odds_ratio": -0.6378474831581116,
|
| 340 |
+
"logps/chosen": -1.2965327501296997,
|
| 341 |
+
"logps/rejected": -3.484023332595825,
|
| 342 |
+
"loss": 40.3825,
|
| 343 |
+
"nll_loss": 1.5268816947937012,
|
| 344 |
+
"po_loss": -0.2649289071559906,
|
| 345 |
+
"rewards/accuracies": 0.671875,
|
| 346 |
+
"rewards/chosen": -0.6482663750648499,
|
| 347 |
+
"rewards/margins": 1.093745231628418,
|
| 348 |
+
"rewards/rejected": -1.7420116662979126,
|
| 349 |
+
"step": 100
|
| 350 |
+
},
|
| 351 |
+
{
|
| 352 |
+
"epoch": 0.896,
|
| 353 |
+
"grad_norm": 120.91621398925781,
|
| 354 |
+
"learning_rate": 4.431042398061499e-05,
|
| 355 |
+
"log_odds_chosen": 2.5239200592041016,
|
| 356 |
+
"log_odds_ratio": -0.7032963633537292,
|
| 357 |
+
"logps/chosen": -2.863506317138672,
|
| 358 |
+
"logps/rejected": -5.296839714050293,
|
| 359 |
+
"loss": 85.1499,
|
| 360 |
+
"nll_loss": 2.635256767272949,
|
| 361 |
+
"po_loss": 0.025676274672150612,
|
| 362 |
+
"rewards/accuracies": 0.6937500238418579,
|
| 363 |
+
"rewards/chosen": -1.431753158569336,
|
| 364 |
+
"rewards/margins": 1.2166664600372314,
|
| 365 |
+
"rewards/rejected": -2.6484198570251465,
|
| 366 |
+
"step": 105
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.9386666666666666,
|
| 370 |
+
"grad_norm": 80.25611114501953,
|
| 371 |
+
"learning_rate": 4.34949719011896e-05,
|
| 372 |
+
"log_odds_chosen": 0.859009861946106,
|
| 373 |
+
"log_odds_ratio": -0.5764500498771667,
|
| 374 |
+
"logps/chosen": -1.229062557220459,
|
| 375 |
+
"logps/rejected": -1.9831807613372803,
|
| 376 |
+
"loss": 47.4796,
|
| 377 |
+
"nll_loss": 1.5179859399795532,
|
| 378 |
+
"po_loss": -0.03424710035324097,
|
| 379 |
+
"rewards/accuracies": 0.6625000238418579,
|
| 380 |
+
"rewards/chosen": -0.6145312786102295,
|
| 381 |
+
"rewards/margins": 0.37705910205841064,
|
| 382 |
+
"rewards/rejected": -0.9915903806686401,
|
| 383 |
+
"step": 110
|
| 384 |
+
},
|
| 385 |
+
{
|
| 386 |
+
"epoch": 0.9813333333333333,
|
| 387 |
+
"grad_norm": 271.3537902832031,
|
| 388 |
+
"learning_rate": 4.263353840751022e-05,
|
| 389 |
+
"log_odds_chosen": 1.4788315296173096,
|
| 390 |
+
"log_odds_ratio": -0.5846825242042542,
|
| 391 |
+
"logps/chosen": -1.1537017822265625,
|
| 392 |
+
"logps/rejected": -2.52790904045105,
|
| 393 |
+
"loss": 44.319,
|
| 394 |
+
"nll_loss": 1.4921709299087524,
|
| 395 |
+
"po_loss": -0.10720314830541611,
|
| 396 |
+
"rewards/accuracies": 0.6781250238418579,
|
| 397 |
+
"rewards/chosen": -0.5768508911132812,
|
| 398 |
+
"rewards/margins": 0.6871034502983093,
|
| 399 |
+
"rewards/rejected": -1.263954520225525,
|
| 400 |
+
"step": 115
|
| 401 |
+
},
|
| 402 |
+
{
|
| 403 |
+
"epoch": 1.0170666666666666,
|
| 404 |
+
"grad_norm": 146.54196166992188,
|
| 405 |
+
"learning_rate": 4.172826515897146e-05,
|
| 406 |
+
"log_odds_chosen": 2.831665515899658,
|
| 407 |
+
"log_odds_ratio": -0.4861010015010834,
|
| 408 |
+
"logps/chosen": -1.089508056640625,
|
| 409 |
+
"logps/rejected": -3.7528676986694336,
|
| 410 |
+
"loss": 33.7801,
|
| 411 |
+
"nll_loss": 1.4263149499893188,
|
| 412 |
+
"po_loss": -0.1658623069524765,
|
| 413 |
+
"rewards/accuracies": 0.7723880410194397,
|
| 414 |
+
"rewards/chosen": -0.5447540283203125,
|
| 415 |
+
"rewards/margins": 1.3316797018051147,
|
| 416 |
+
"rewards/rejected": -1.8764338493347168,
|
| 417 |
+
"step": 120
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 1.0597333333333334,
|
| 421 |
+
"grad_norm": 1590.6824951171875,
|
| 422 |
+
"learning_rate": 4.078140280750597e-05,
|
| 423 |
+
"log_odds_chosen": 4.709661483764648,
|
| 424 |
+
"log_odds_ratio": -0.41537848114967346,
|
| 425 |
+
"logps/chosen": -1.5830024480819702,
|
| 426 |
+
"logps/rejected": -6.075448036193848,
|
| 427 |
+
"loss": 46.6655,
|
| 428 |
+
"nll_loss": 1.6417958736419678,
|
| 429 |
+
"po_loss": -0.18349842727184296,
|
| 430 |
+
"rewards/accuracies": 0.8531249761581421,
|
| 431 |
+
"rewards/chosen": -0.7915012240409851,
|
| 432 |
+
"rewards/margins": 2.246222496032715,
|
| 433 |
+
"rewards/rejected": -3.037724018096924,
|
| 434 |
+
"step": 125
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 1.1024,
|
| 438 |
+
"grad_norm": 321.93255615234375,
|
| 439 |
+
"learning_rate": 3.9795305402109195e-05,
|
| 440 |
+
"log_odds_chosen": 3.6606578826904297,
|
| 441 |
+
"log_odds_ratio": -0.4261302947998047,
|
| 442 |
+
"logps/chosen": -1.2628165483474731,
|
| 443 |
+
"logps/rejected": -4.708868980407715,
|
| 444 |
+
"loss": 36.688,
|
| 445 |
+
"nll_loss": 1.4487800598144531,
|
| 446 |
+
"po_loss": -0.3022807240486145,
|
| 447 |
+
"rewards/accuracies": 0.8218749761581421,
|
| 448 |
+
"rewards/chosen": -0.6314082741737366,
|
| 449 |
+
"rewards/margins": 1.7230265140533447,
|
| 450 |
+
"rewards/rejected": -2.3544344902038574,
|
| 451 |
+
"step": 130
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 1.1450666666666667,
|
| 455 |
+
"grad_norm": 156.33656311035156,
|
| 456 |
+
"learning_rate": 3.8772424536302564e-05,
|
| 457 |
+
"log_odds_chosen": 3.3216488361358643,
|
| 458 |
+
"log_odds_ratio": -0.45871955156326294,
|
| 459 |
+
"logps/chosen": -1.054512619972229,
|
| 460 |
+
"logps/rejected": -4.166798114776611,
|
| 461 |
+
"loss": 37.5147,
|
| 462 |
+
"nll_loss": 1.3529837131500244,
|
| 463 |
+
"po_loss": -0.18065081536769867,
|
| 464 |
+
"rewards/accuracies": 0.778124988079071,
|
| 465 |
+
"rewards/chosen": -0.5272563099861145,
|
| 466 |
+
"rewards/margins": 1.5561425685882568,
|
| 467 |
+
"rewards/rejected": -2.0833990573883057,
|
| 468 |
+
"step": 135
|
| 469 |
+
},
|
| 470 |
+
{
|
| 471 |
+
"epoch": 1.1877333333333333,
|
| 472 |
+
"grad_norm": 126.58124542236328,
|
| 473 |
+
"learning_rate": 3.771530325308579e-05,
|
| 474 |
+
"log_odds_chosen": 3.002664089202881,
|
| 475 |
+
"log_odds_ratio": -0.3991411030292511,
|
| 476 |
+
"logps/chosen": -0.99451744556427,
|
| 477 |
+
"logps/rejected": -3.7320716381073,
|
| 478 |
+
"loss": 33.4168,
|
| 479 |
+
"nll_loss": 1.2971866130828857,
|
| 480 |
+
"po_loss": -0.2529115378856659,
|
| 481 |
+
"rewards/accuracies": 0.862500011920929,
|
| 482 |
+
"rewards/chosen": -0.497258722782135,
|
| 483 |
+
"rewards/margins": 1.3687771558761597,
|
| 484 |
+
"rewards/rejected": -1.86603581905365,
|
| 485 |
+
"step": 140
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 1.2304,
|
| 489 |
+
"grad_norm": 144.6908416748047,
|
| 490 |
+
"learning_rate": 3.662656972253127e-05,
|
| 491 |
+
"log_odds_chosen": 3.620790958404541,
|
| 492 |
+
"log_odds_ratio": -0.39333289861679077,
|
| 493 |
+
"logps/chosen": -1.1291520595550537,
|
| 494 |
+
"logps/rejected": -4.468079090118408,
|
| 495 |
+
"loss": 31.6104,
|
| 496 |
+
"nll_loss": 1.2871015071868896,
|
| 497 |
+
"po_loss": -0.2992752194404602,
|
| 498 |
+
"rewards/accuracies": 0.875,
|
| 499 |
+
"rewards/chosen": -0.5645760297775269,
|
| 500 |
+
"rewards/margins": 1.6694637537002563,
|
| 501 |
+
"rewards/rejected": -2.234039545059204,
|
| 502 |
+
"step": 145
|
| 503 |
+
},
|
| 504 |
+
{
|
| 505 |
+
"epoch": 1.2730666666666668,
|
| 506 |
+
"grad_norm": 317.3659362792969,
|
| 507 |
+
"learning_rate": 3.550893070773914e-05,
|
| 508 |
+
"log_odds_chosen": 2.5045361518859863,
|
| 509 |
+
"log_odds_ratio": -0.5062359571456909,
|
| 510 |
+
"logps/chosen": -1.01883864402771,
|
| 511 |
+
"logps/rejected": -3.3051059246063232,
|
| 512 |
+
"loss": 35.2667,
|
| 513 |
+
"nll_loss": 1.3030840158462524,
|
| 514 |
+
"po_loss": -0.20100060105323792,
|
| 515 |
+
"rewards/accuracies": 0.831250011920929,
|
| 516 |
+
"rewards/chosen": -0.509419322013855,
|
| 517 |
+
"rewards/margins": 1.143133521080017,
|
| 518 |
+
"rewards/rejected": -1.6525529623031616,
|
| 519 |
+
"step": 150
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 1.3157333333333332,
|
| 523 |
+
"grad_norm": 99.71307373046875,
|
| 524 |
+
"learning_rate": 3.436516483539781e-05,
|
| 525 |
+
"log_odds_chosen": 2.3835015296936035,
|
| 526 |
+
"log_odds_ratio": -0.4490797519683838,
|
| 527 |
+
"logps/chosen": -0.961607813835144,
|
| 528 |
+
"logps/rejected": -3.094728946685791,
|
| 529 |
+
"loss": 36.1188,
|
| 530 |
+
"nll_loss": 1.2786829471588135,
|
| 531 |
+
"po_loss": -0.1499709188938141,
|
| 532 |
+
"rewards/accuracies": 0.8343750238418579,
|
| 533 |
+
"rewards/chosen": -0.480803906917572,
|
| 534 |
+
"rewards/margins": 1.0665608644485474,
|
| 535 |
+
"rewards/rejected": -1.5473644733428955,
|
| 536 |
+
"step": 155
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 1.3584,
|
| 540 |
+
"grad_norm": 176.96144104003906,
|
| 541 |
+
"learning_rate": 3.3198115687680115e-05,
|
| 542 |
+
"log_odds_chosen": 2.804508924484253,
|
| 543 |
+
"log_odds_ratio": -0.4551823139190674,
|
| 544 |
+
"logps/chosen": -0.9649540781974792,
|
| 545 |
+
"logps/rejected": -3.5283005237579346,
|
| 546 |
+
"loss": 31.9084,
|
| 547 |
+
"nll_loss": 1.2769290208816528,
|
| 548 |
+
"po_loss": -0.2797916531562805,
|
| 549 |
+
"rewards/accuracies": 0.7906249761581421,
|
| 550 |
+
"rewards/chosen": -0.4824770390987396,
|
| 551 |
+
"rewards/margins": 1.2816731929779053,
|
| 552 |
+
"rewards/rejected": -1.7641502618789673,
|
| 553 |
+
"step": 160
|
| 554 |
+
},
|
| 555 |
+
{
|
| 556 |
+
"epoch": 1.4010666666666667,
|
| 557 |
+
"grad_norm": 86.9504165649414,
|
| 558 |
+
"learning_rate": 3.201068473265007e-05,
|
| 559 |
+
"log_odds_chosen": 4.039158821105957,
|
| 560 |
+
"log_odds_ratio": -0.41200175881385803,
|
| 561 |
+
"logps/chosen": -1.1530981063842773,
|
| 562 |
+
"logps/rejected": -4.936544418334961,
|
| 563 |
+
"loss": 33.9241,
|
| 564 |
+
"nll_loss": 1.339659571647644,
|
| 565 |
+
"po_loss": -0.27953004837036133,
|
| 566 |
+
"rewards/accuracies": 0.840624988079071,
|
| 567 |
+
"rewards/chosen": -0.5765490531921387,
|
| 568 |
+
"rewards/margins": 1.8917229175567627,
|
| 569 |
+
"rewards/rejected": -2.4682722091674805,
|
| 570 |
+
"step": 165
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 1.4437333333333333,
|
| 574 |
+
"grad_norm": 67.8962631225586,
|
| 575 |
+
"learning_rate": 3.0805824110756064e-05,
|
| 576 |
+
"log_odds_chosen": 2.8494575023651123,
|
| 577 |
+
"log_odds_ratio": -0.4309585690498352,
|
| 578 |
+
"logps/chosen": -1.0003385543823242,
|
| 579 |
+
"logps/rejected": -3.5997538566589355,
|
| 580 |
+
"loss": 32.9378,
|
| 581 |
+
"nll_loss": 1.320733904838562,
|
| 582 |
+
"po_loss": -0.2914286255836487,
|
| 583 |
+
"rewards/accuracies": 0.84375,
|
| 584 |
+
"rewards/chosen": -0.5001692771911621,
|
| 585 |
+
"rewards/margins": 1.2997074127197266,
|
| 586 |
+
"rewards/rejected": -1.7998769283294678,
|
| 587 |
+
"step": 170
|
| 588 |
+
},
|
| 589 |
+
{
|
| 590 |
+
"epoch": 1.4864,
|
| 591 |
+
"grad_norm": 332.5802307128906,
|
| 592 |
+
"learning_rate": 2.958652929534456e-05,
|
| 593 |
+
"log_odds_chosen": 4.219790458679199,
|
| 594 |
+
"log_odds_ratio": -0.3872655928134918,
|
| 595 |
+
"logps/chosen": -0.9488153457641602,
|
| 596 |
+
"logps/rejected": -4.870639324188232,
|
| 597 |
+
"loss": 24.5,
|
| 598 |
+
"nll_loss": 1.2430005073547363,
|
| 599 |
+
"po_loss": -0.47737568616867065,
|
| 600 |
+
"rewards/accuracies": 0.8656250238418579,
|
| 601 |
+
"rewards/chosen": -0.4744076728820801,
|
| 602 |
+
"rewards/margins": 1.9609121084213257,
|
| 603 |
+
"rewards/rejected": -2.435319662094116,
|
| 604 |
+
"step": 175
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 1.5290666666666666,
|
| 608 |
+
"grad_norm": 120.37972259521484,
|
| 609 |
+
"learning_rate": 2.8355831645441388e-05,
|
| 610 |
+
"log_odds_chosen": 3.8075263500213623,
|
| 611 |
+
"log_odds_ratio": -0.4285067617893219,
|
| 612 |
+
"logps/chosen": -1.0132713317871094,
|
| 613 |
+
"logps/rejected": -4.545207500457764,
|
| 614 |
+
"loss": 30.9088,
|
| 615 |
+
"nll_loss": 1.2752387523651123,
|
| 616 |
+
"po_loss": -0.3093392252922058,
|
| 617 |
+
"rewards/accuracies": 0.840624988079071,
|
| 618 |
+
"rewards/chosen": -0.5066356658935547,
|
| 619 |
+
"rewards/margins": 1.7659679651260376,
|
| 620 |
+
"rewards/rejected": -2.272603750228882,
|
| 621 |
+
"step": 180
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"epoch": 1.5717333333333334,
|
| 625 |
+
"grad_norm": 74.23609924316406,
|
| 626 |
+
"learning_rate": 2.7116790869315582e-05,
|
| 627 |
+
"log_odds_chosen": 3.5458972454071045,
|
| 628 |
+
"log_odds_ratio": -0.3724256157875061,
|
| 629 |
+
"logps/chosen": -0.9797577857971191,
|
| 630 |
+
"logps/rejected": -4.209129810333252,
|
| 631 |
+
"loss": 27.577,
|
| 632 |
+
"nll_loss": 1.2894318103790283,
|
| 633 |
+
"po_loss": -0.42764997482299805,
|
| 634 |
+
"rewards/accuracies": 0.887499988079071,
|
| 635 |
+
"rewards/chosen": -0.48987889289855957,
|
| 636 |
+
"rewards/margins": 1.6146857738494873,
|
| 637 |
+
"rewards/rejected": -2.104564905166626,
|
| 638 |
+
"step": 185
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 1.6143999999999998,
|
| 642 |
+
"grad_norm": 1912.4134521484375,
|
| 643 |
+
"learning_rate": 2.587248741756253e-05,
|
| 644 |
+
"log_odds_chosen": 4.569626808166504,
|
| 645 |
+
"log_odds_ratio": -0.6492961049079895,
|
| 646 |
+
"logps/chosen": -1.6978082656860352,
|
| 647 |
+
"logps/rejected": -5.967839241027832,
|
| 648 |
+
"loss": 47.8645,
|
| 649 |
+
"nll_loss": 1.7877342700958252,
|
| 650 |
+
"po_loss": -0.2919684648513794,
|
| 651 |
+
"rewards/accuracies": 0.840624988079071,
|
| 652 |
+
"rewards/chosen": -0.8489041328430176,
|
| 653 |
+
"rewards/margins": 2.1350154876708984,
|
| 654 |
+
"rewards/rejected": -2.983919620513916,
|
| 655 |
+
"step": 190
|
| 656 |
+
},
|
| 657 |
+
{
|
| 658 |
+
"epoch": 1.6570666666666667,
|
| 659 |
+
"grad_norm": 246.9727020263672,
|
| 660 |
+
"learning_rate": 2.4626014824618415e-05,
|
| 661 |
+
"log_odds_chosen": 3.452259063720703,
|
| 662 |
+
"log_odds_ratio": -0.4340329170227051,
|
| 663 |
+
"logps/chosen": -1.348034143447876,
|
| 664 |
+
"logps/rejected": -4.516729831695557,
|
| 665 |
+
"loss": 39.5128,
|
| 666 |
+
"nll_loss": 1.5523712635040283,
|
| 667 |
+
"po_loss": -0.3175976276397705,
|
| 668 |
+
"rewards/accuracies": 0.8218749761581421,
|
| 669 |
+
"rewards/chosen": -0.674017071723938,
|
| 670 |
+
"rewards/margins": 1.5843479633331299,
|
| 671 |
+
"rewards/rejected": -2.2583649158477783,
|
| 672 |
+
"step": 195
|
| 673 |
+
},
|
| 674 |
+
{
|
| 675 |
+
"epoch": 1.6997333333333333,
|
| 676 |
+
"grad_norm": 249.69097900390625,
|
| 677 |
+
"learning_rate": 2.3380472017746202e-05,
|
| 678 |
+
"log_odds_chosen": 4.080594539642334,
|
| 679 |
+
"log_odds_ratio": -0.41809386014938354,
|
| 680 |
+
"logps/chosen": -1.1225640773773193,
|
| 681 |
+
"logps/rejected": -4.919617652893066,
|
| 682 |
+
"loss": 22.4016,
|
| 683 |
+
"nll_loss": 1.3972797393798828,
|
| 684 |
+
"po_loss": -0.6972296237945557,
|
| 685 |
+
"rewards/accuracies": 0.862500011920929,
|
| 686 |
+
"rewards/chosen": -0.5612820386886597,
|
| 687 |
+
"rewards/margins": 1.8985265493392944,
|
| 688 |
+
"rewards/rejected": -2.459808826446533,
|
| 689 |
+
"step": 200
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 1.7424,
|
| 693 |
+
"grad_norm": 70.95199584960938,
|
| 694 |
+
"learning_rate": 2.2138955612614207e-05,
|
| 695 |
+
"log_odds_chosen": 3.4808452129364014,
|
| 696 |
+
"log_odds_ratio": -0.4113942086696625,
|
| 697 |
+
"logps/chosen": -0.9912225604057312,
|
| 698 |
+
"logps/rejected": -4.186377048492432,
|
| 699 |
+
"loss": 30.6187,
|
| 700 |
+
"nll_loss": 1.2635384798049927,
|
| 701 |
+
"po_loss": -0.30670270323753357,
|
| 702 |
+
"rewards/accuracies": 0.862500011920929,
|
| 703 |
+
"rewards/chosen": -0.4956112802028656,
|
| 704 |
+
"rewards/margins": 1.5975773334503174,
|
| 705 |
+
"rewards/rejected": -2.093188524246216,
|
| 706 |
+
"step": 205
|
| 707 |
+
},
|
| 708 |
+
{
|
| 709 |
+
"epoch": 1.7850666666666668,
|
| 710 |
+
"grad_norm": 77.6159896850586,
|
| 711 |
+
"learning_rate": 2.090455221462156e-05,
|
| 712 |
+
"log_odds_chosen": 3.126645803451538,
|
| 713 |
+
"log_odds_ratio": -0.4194856286048889,
|
| 714 |
+
"logps/chosen": -0.976054847240448,
|
| 715 |
+
"logps/rejected": -3.822575807571411,
|
| 716 |
+
"loss": 30.5635,
|
| 717 |
+
"nll_loss": 1.2749359607696533,
|
| 718 |
+
"po_loss": -0.319825679063797,
|
| 719 |
+
"rewards/accuracies": 0.8500000238418579,
|
| 720 |
+
"rewards/chosen": -0.488027423620224,
|
| 721 |
+
"rewards/margins": 1.4232604503631592,
|
| 722 |
+
"rewards/rejected": -1.9112879037857056,
|
| 723 |
+
"step": 210
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 1.8277333333333332,
|
| 727 |
+
"grad_norm": 191.34779357910156,
|
| 728 |
+
"learning_rate": 1.9680330745110954e-05,
|
| 729 |
+
"log_odds_chosen": 4.950188636779785,
|
| 730 |
+
"log_odds_ratio": -0.3956086039543152,
|
| 731 |
+
"logps/chosen": -0.9748937487602234,
|
| 732 |
+
"logps/rejected": -5.643929481506348,
|
| 733 |
+
"loss": 23.1695,
|
| 734 |
+
"nll_loss": 1.2857718467712402,
|
| 735 |
+
"po_loss": -0.5617233514785767,
|
| 736 |
+
"rewards/accuracies": 0.84375,
|
| 737 |
+
"rewards/chosen": -0.4874468743801117,
|
| 738 |
+
"rewards/margins": 2.33451771736145,
|
| 739 |
+
"rewards/rejected": -2.821964740753174,
|
| 740 |
+
"step": 215
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 1.8704,
|
| 744 |
+
"grad_norm": 327.7476501464844,
|
| 745 |
+
"learning_rate": 1.8469334811546542e-05,
|
| 746 |
+
"log_odds_chosen": 4.540267467498779,
|
| 747 |
+
"log_odds_ratio": -0.39753258228302,
|
| 748 |
+
"logps/chosen": -0.9387453198432922,
|
| 749 |
+
"logps/rejected": -5.1981964111328125,
|
| 750 |
+
"loss": 28.4212,
|
| 751 |
+
"nll_loss": 1.252907633781433,
|
| 752 |
+
"po_loss": -0.36474356055259705,
|
| 753 |
+
"rewards/accuracies": 0.859375,
|
| 754 |
+
"rewards/chosen": -0.4693726599216461,
|
| 755 |
+
"rewards/margins": 2.129725217819214,
|
| 756 |
+
"rewards/rejected": -2.5990982055664062,
|
| 757 |
+
"step": 220
|
| 758 |
+
},
|
| 759 |
+
{
|
| 760 |
+
"epoch": 1.9130666666666667,
|
| 761 |
+
"grad_norm": 540.2483520507812,
|
| 762 |
+
"learning_rate": 1.7274575140626318e-05,
|
| 763 |
+
"log_odds_chosen": 5.64566707611084,
|
| 764 |
+
"log_odds_ratio": -0.4446595311164856,
|
| 765 |
+
"logps/chosen": -1.22698974609375,
|
| 766 |
+
"logps/rejected": -6.585175514221191,
|
| 767 |
+
"loss": 34.7211,
|
| 768 |
+
"nll_loss": 1.3462177515029907,
|
| 769 |
+
"po_loss": -0.26118165254592896,
|
| 770 |
+
"rewards/accuracies": 0.8656250238418579,
|
| 771 |
+
"rewards/chosen": -0.613494873046875,
|
| 772 |
+
"rewards/margins": 2.6790928840637207,
|
| 773 |
+
"rewards/rejected": -3.2925877571105957,
|
| 774 |
+
"step": 225
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 1.9557333333333333,
|
| 778 |
+
"grad_norm": 401.98248291015625,
|
| 779 |
+
"learning_rate": 1.609902209314108e-05,
|
| 780 |
+
"log_odds_chosen": 4.817292213439941,
|
| 781 |
+
"log_odds_ratio": -0.42455047369003296,
|
| 782 |
+
"logps/chosen": -1.0640500783920288,
|
| 783 |
+
"logps/rejected": -5.626768589019775,
|
| 784 |
+
"loss": 31.062,
|
| 785 |
+
"nll_loss": 1.342284917831421,
|
| 786 |
+
"po_loss": -0.37159785628318787,
|
| 787 |
+
"rewards/accuracies": 0.8187500238418579,
|
| 788 |
+
"rewards/chosen": -0.5320250391960144,
|
| 789 |
+
"rewards/margins": 2.2813591957092285,
|
| 790 |
+
"rewards/rejected": -2.8133842945098877,
|
| 791 |
+
"step": 230
|
| 792 |
+
},
|
| 793 |
+
{
|
| 794 |
+
"epoch": 1.9984,
|
| 795 |
+
"grad_norm": 91.74356079101562,
|
| 796 |
+
"learning_rate": 1.4945598279189565e-05,
|
| 797 |
+
"log_odds_chosen": 5.049031734466553,
|
| 798 |
+
"log_odds_ratio": -0.406465619802475,
|
| 799 |
+
"logps/chosen": -0.9556134939193726,
|
| 800 |
+
"logps/rejected": -5.697612762451172,
|
| 801 |
+
"loss": 25.4086,
|
| 802 |
+
"nll_loss": 1.2838685512542725,
|
| 803 |
+
"po_loss": -0.48984843492507935,
|
| 804 |
+
"rewards/accuracies": 0.840624988079071,
|
| 805 |
+
"rewards/chosen": -0.4778067469596863,
|
| 806 |
+
"rewards/margins": 2.370999813079834,
|
| 807 |
+
"rewards/rejected": -2.848806381225586,
|
| 808 |
+
"step": 235
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 2.034133333333333,
|
| 812 |
+
"grad_norm": 84.14781188964844,
|
| 813 |
+
"learning_rate": 1.3817171292109183e-05,
|
| 814 |
+
"log_odds_chosen": 4.798027992248535,
|
| 815 |
+
"log_odds_ratio": -0.3244837820529938,
|
| 816 |
+
"logps/chosen": -0.8924720287322998,
|
| 817 |
+
"logps/rejected": -5.292112827301025,
|
| 818 |
+
"loss": 14.7486,
|
| 819 |
+
"nll_loss": 1.1619552373886108,
|
| 820 |
+
"po_loss": -0.6116347908973694,
|
| 821 |
+
"rewards/accuracies": 0.9365671873092651,
|
| 822 |
+
"rewards/chosen": -0.4462360143661499,
|
| 823 |
+
"rewards/margins": 2.1998205184936523,
|
| 824 |
+
"rewards/rejected": -2.6460564136505127,
|
| 825 |
+
"step": 240
|
| 826 |
+
},
|
| 827 |
+
{
|
| 828 |
+
"epoch": 2.0768,
|
| 829 |
+
"grad_norm": 265.51214599609375,
|
| 830 |
+
"learning_rate": 1.271654657918722e-05,
|
| 831 |
+
"log_odds_chosen": 3.9462997913360596,
|
| 832 |
+
"log_odds_ratio": -0.29830431938171387,
|
| 833 |
+
"logps/chosen": -0.8560145497322083,
|
| 834 |
+
"logps/rejected": -4.366959571838379,
|
| 835 |
+
"loss": 21.2586,
|
| 836 |
+
"nll_loss": 1.1650712490081787,
|
| 837 |
+
"po_loss": -0.5007405281066895,
|
| 838 |
+
"rewards/accuracies": 0.9437500238418579,
|
| 839 |
+
"rewards/chosen": -0.4280072748661041,
|
| 840 |
+
"rewards/margins": 1.7554724216461182,
|
| 841 |
+
"rewards/rejected": -2.1834797859191895,
|
| 842 |
+
"step": 245
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 2.119466666666667,
|
| 846 |
+
"grad_norm": 189.54098510742188,
|
| 847 |
+
"learning_rate": 1.1646460466876783e-05,
|
| 848 |
+
"log_odds_chosen": 5.587141036987305,
|
| 849 |
+
"log_odds_ratio": -0.23089304566383362,
|
| 850 |
+
"logps/chosen": -0.941892147064209,
|
| 851 |
+
"logps/rejected": -6.023226737976074,
|
| 852 |
+
"loss": 15.3802,
|
| 853 |
+
"nll_loss": 1.1797395944595337,
|
| 854 |
+
"po_loss": -0.6991080641746521,
|
| 855 |
+
"rewards/accuracies": 0.9437500238418579,
|
| 856 |
+
"rewards/chosen": -0.4709460735321045,
|
| 857 |
+
"rewards/margins": 2.5406672954559326,
|
| 858 |
+
"rewards/rejected": -3.011613368988037,
|
| 859 |
+
"step": 250
|
| 860 |
+
},
|
| 861 |
+
{
|
| 862 |
+
"epoch": 2.1621333333333332,
|
| 863 |
+
"grad_norm": 198.67486572265625,
|
| 864 |
+
"learning_rate": 1.0609573357858166e-05,
|
| 865 |
+
"log_odds_chosen": 6.2472615242004395,
|
| 866 |
+
"log_odds_ratio": -0.2708757519721985,
|
| 867 |
+
"logps/chosen": -1.0158169269561768,
|
| 868 |
+
"logps/rejected": -6.834556579589844,
|
| 869 |
+
"loss": 14.1929,
|
| 870 |
+
"nll_loss": 1.237670660018921,
|
| 871 |
+
"po_loss": -0.7941441535949707,
|
| 872 |
+
"rewards/accuracies": 0.925000011920929,
|
| 873 |
+
"rewards/chosen": -0.5079084634780884,
|
| 874 |
+
"rewards/margins": 2.909369945526123,
|
| 875 |
+
"rewards/rejected": -3.417278289794922,
|
| 876 |
+
"step": 255
|
| 877 |
+
},
|
| 878 |
+
{
|
| 879 |
+
"epoch": 2.2048,
|
| 880 |
+
"grad_norm": 758.944580078125,
|
| 881 |
+
"learning_rate": 9.608463116858542e-06,
|
| 882 |
+
"log_odds_chosen": 6.66796350479126,
|
| 883 |
+
"log_odds_ratio": -0.23688967525959015,
|
| 884 |
+
"logps/chosen": -0.9843616485595703,
|
| 885 |
+
"logps/rejected": -7.208956718444824,
|
| 886 |
+
"loss": 11.9915,
|
| 887 |
+
"nll_loss": 1.2095975875854492,
|
| 888 |
+
"po_loss": -0.8348617553710938,
|
| 889 |
+
"rewards/accuracies": 0.9593750238418579,
|
| 890 |
+
"rewards/chosen": -0.49218082427978516,
|
| 891 |
+
"rewards/margins": 3.112297534942627,
|
| 892 |
+
"rewards/rejected": -3.604478359222412,
|
| 893 |
+
"step": 260
|
| 894 |
+
},
|
| 895 |
+
{
|
| 896 |
+
"epoch": 2.2474666666666665,
|
| 897 |
+
"grad_norm": 504.9031982421875,
|
| 898 |
+
"learning_rate": 8.645618661674142e-06,
|
| 899 |
+
"log_odds_chosen": 6.688530921936035,
|
| 900 |
+
"log_odds_ratio": -0.2643258273601532,
|
| 901 |
+
"logps/chosen": -1.105387806892395,
|
| 902 |
+
"logps/rejected": -7.340773582458496,
|
| 903 |
+
"loss": 11.084,
|
| 904 |
+
"nll_loss": 1.2117825746536255,
|
| 905 |
+
"po_loss": -0.8654062151908875,
|
| 906 |
+
"rewards/accuracies": 0.953125,
|
| 907 |
+
"rewards/chosen": -0.5526939034461975,
|
| 908 |
+
"rewards/margins": 3.1176929473876953,
|
| 909 |
+
"rewards/rejected": -3.670386791229248,
|
| 910 |
+
"step": 265
|
| 911 |
+
},
|
| 912 |
+
{
|
| 913 |
+
"epoch": 2.2901333333333334,
|
| 914 |
+
"grad_norm": 250.27699279785156,
|
| 915 |
+
"learning_rate": 7.723433775328384e-06,
|
| 916 |
+
"log_odds_chosen": 6.858447074890137,
|
| 917 |
+
"log_odds_ratio": -0.31602945923805237,
|
| 918 |
+
"logps/chosen": -1.1505261659622192,
|
| 919 |
+
"logps/rejected": -7.580657005310059,
|
| 920 |
+
"loss": 20.3187,
|
| 921 |
+
"nll_loss": 1.2361539602279663,
|
| 922 |
+
"po_loss": -0.6011937260627747,
|
| 923 |
+
"rewards/accuracies": 0.925000011920929,
|
| 924 |
+
"rewards/chosen": -0.5752630829811096,
|
| 925 |
+
"rewards/margins": 3.2150657176971436,
|
| 926 |
+
"rewards/rejected": -3.7903285026550293,
|
| 927 |
+
"step": 270
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 2.3327999999999998,
|
| 931 |
+
"grad_norm": 3240.22802734375,
|
| 932 |
+
"learning_rate": 6.844201154750177e-06,
|
| 933 |
+
"log_odds_chosen": 6.9127960205078125,
|
| 934 |
+
"log_odds_ratio": -0.28268226981163025,
|
| 935 |
+
"logps/chosen": -1.3622684478759766,
|
| 936 |
+
"logps/rejected": -7.855978965759277,
|
| 937 |
+
"loss": 22.8393,
|
| 938 |
+
"nll_loss": 1.4035627841949463,
|
| 939 |
+
"po_loss": -0.6898348927497864,
|
| 940 |
+
"rewards/accuracies": 0.9437500238418579,
|
| 941 |
+
"rewards/chosen": -0.6811342239379883,
|
| 942 |
+
"rewards/margins": 3.2468554973602295,
|
| 943 |
+
"rewards/rejected": -3.9279894828796387,
|
| 944 |
+
"step": 275
|
| 945 |
+
},
|
| 946 |
+
{
|
| 947 |
+
"epoch": 2.3754666666666666,
|
| 948 |
+
"grad_norm": 650.8209228515625,
|
| 949 |
+
"learning_rate": 6.010106710768052e-06,
|
| 950 |
+
"log_odds_chosen": 6.493264675140381,
|
| 951 |
+
"log_odds_ratio": -0.34161344170570374,
|
| 952 |
+
"logps/chosen": -1.386515498161316,
|
| 953 |
+
"logps/rejected": -7.452689170837402,
|
| 954 |
+
"loss": 19.5164,
|
| 955 |
+
"nll_loss": 1.3803019523620605,
|
| 956 |
+
"po_loss": -0.7704136967658997,
|
| 957 |
+
"rewards/accuracies": 0.90625,
|
| 958 |
+
"rewards/chosen": -0.693257749080658,
|
| 959 |
+
"rewards/margins": 3.0330867767333984,
|
| 960 |
+
"rewards/rejected": -3.726344585418701,
|
| 961 |
+
"step": 280
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 2.4181333333333335,
|
| 965 |
+
"grad_norm": 507.415283203125,
|
| 966 |
+
"learning_rate": 5.223224133591476e-06,
|
| 967 |
+
"log_odds_chosen": 6.825043678283691,
|
| 968 |
+
"log_odds_ratio": -0.2423882931470871,
|
| 969 |
+
"logps/chosen": -1.1171207427978516,
|
| 970 |
+
"logps/rejected": -7.483119964599609,
|
| 971 |
+
"loss": 19.53,
|
| 972 |
+
"nll_loss": 1.2847764492034912,
|
| 973 |
+
"po_loss": -0.6744629144668579,
|
| 974 |
+
"rewards/accuracies": 0.9468749761581421,
|
| 975 |
+
"rewards/chosen": -0.5585603713989258,
|
| 976 |
+
"rewards/margins": 3.1829993724823,
|
| 977 |
+
"rewards/rejected": -3.7415599822998047,
|
| 978 |
+
"step": 285
|
| 979 |
+
},
|
| 980 |
+
{
|
| 981 |
+
"epoch": 2.4608,
|
| 982 |
+
"grad_norm": 789.3963623046875,
|
| 983 |
+
"learning_rate": 4.4855097372902135e-06,
|
| 984 |
+
"log_odds_chosen": 7.850038051605225,
|
| 985 |
+
"log_odds_ratio": -0.26526254415512085,
|
| 986 |
+
"logps/chosen": -1.4054704904556274,
|
| 987 |
+
"logps/rejected": -8.81872844696045,
|
| 988 |
+
"loss": 18.6059,
|
| 989 |
+
"nll_loss": 1.3957431316375732,
|
| 990 |
+
"po_loss": -0.8143098950386047,
|
| 991 |
+
"rewards/accuracies": 0.9312499761581421,
|
| 992 |
+
"rewards/chosen": -0.7027352452278137,
|
| 993 |
+
"rewards/margins": 3.7066292762756348,
|
| 994 |
+
"rewards/rejected": -4.409364223480225,
|
| 995 |
+
"step": 290
|
| 996 |
+
},
|
| 997 |
+
{
|
| 998 |
+
"epoch": 2.5034666666666667,
|
| 999 |
+
"grad_norm": 1187.58837890625,
|
| 1000 |
+
"learning_rate": 3.798797596089351e-06,
|
| 1001 |
+
"log_odds_chosen": 8.998720169067383,
|
| 1002 |
+
"log_odds_ratio": -0.218130424618721,
|
| 1003 |
+
"logps/chosen": -1.247434377670288,
|
| 1004 |
+
"logps/rejected": -9.765419960021973,
|
| 1005 |
+
"loss": 10.3413,
|
| 1006 |
+
"nll_loss": 1.355464220046997,
|
| 1007 |
+
"po_loss": -1.032299280166626,
|
| 1008 |
+
"rewards/accuracies": 0.9593750238418579,
|
| 1009 |
+
"rewards/chosen": -0.623717188835144,
|
| 1010 |
+
"rewards/margins": 4.2589921951293945,
|
| 1011 |
+
"rewards/rejected": -4.882709980010986,
|
| 1012 |
+
"step": 295
|
| 1013 |
+
},
|
| 1014 |
+
{
|
| 1015 |
+
"epoch": 2.5461333333333336,
|
| 1016 |
+
"grad_norm": 593.267578125,
|
| 1017 |
+
"learning_rate": 3.164794984571759e-06,
|
| 1018 |
+
"log_odds_chosen": 6.977963924407959,
|
| 1019 |
+
"log_odds_ratio": -0.3458719849586487,
|
| 1020 |
+
"logps/chosen": -1.6929105520248413,
|
| 1021 |
+
"logps/rejected": -8.215627670288086,
|
| 1022 |
+
"loss": 27.4312,
|
| 1023 |
+
"nll_loss": 1.5623621940612793,
|
| 1024 |
+
"po_loss": -0.7051381468772888,
|
| 1025 |
+
"rewards/accuracies": 0.921875,
|
| 1026 |
+
"rewards/chosen": -0.8464552760124207,
|
| 1027 |
+
"rewards/margins": 3.2613587379455566,
|
| 1028 |
+
"rewards/rejected": -4.107813835144043,
|
| 1029 |
+
"step": 300
|
| 1030 |
+
},
|
| 1031 |
+
{
|
| 1032 |
+
"epoch": 2.5888,
|
| 1033 |
+
"grad_norm": 938.1863403320312,
|
| 1034 |
+
"learning_rate": 2.58507813312448e-06,
|
| 1035 |
+
"log_odds_chosen": 6.914087772369385,
|
| 1036 |
+
"log_odds_ratio": -0.28258031606674194,
|
| 1037 |
+
"logps/chosen": -1.2223470211029053,
|
| 1038 |
+
"logps/rejected": -7.694061279296875,
|
| 1039 |
+
"loss": 21.9591,
|
| 1040 |
+
"nll_loss": 1.3453009128570557,
|
| 1041 |
+
"po_loss": -0.6590788960456848,
|
| 1042 |
+
"rewards/accuracies": 0.909375011920929,
|
| 1043 |
+
"rewards/chosen": -0.6111735105514526,
|
| 1044 |
+
"rewards/margins": 3.2358570098876953,
|
| 1045 |
+
"rewards/rejected": -3.8470306396484375,
|
| 1046 |
+
"step": 305
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 2.6314666666666664,
|
| 1050 |
+
"grad_norm": 221.02281188964844,
|
| 1051 |
+
"learning_rate": 2.0610883091816525e-06,
|
| 1052 |
+
"log_odds_chosen": 5.710585594177246,
|
| 1053 |
+
"log_odds_ratio": -0.2850475609302521,
|
| 1054 |
+
"logps/chosen": -1.021317720413208,
|
| 1055 |
+
"logps/rejected": -6.243259906768799,
|
| 1056 |
+
"loss": 16.8653,
|
| 1057 |
+
"nll_loss": 1.247992753982544,
|
| 1058 |
+
"po_loss": -0.7209508419036865,
|
| 1059 |
+
"rewards/accuracies": 0.9281250238418579,
|
| 1060 |
+
"rewards/chosen": -0.510658860206604,
|
| 1061 |
+
"rewards/margins": 2.610970973968506,
|
| 1062 |
+
"rewards/rejected": -3.1216299533843994,
|
| 1063 |
+
"step": 310
|
| 1064 |
+
},
|
| 1065 |
+
{
|
| 1066 |
+
"epoch": 2.6741333333333333,
|
| 1067 |
+
"grad_norm": 475.2201843261719,
|
| 1068 |
+
"learning_rate": 1.59412823400657e-06,
|
| 1069 |
+
"log_odds_chosen": 5.025318145751953,
|
| 1070 |
+
"log_odds_ratio": -0.25998514890670776,
|
| 1071 |
+
"logps/chosen": -1.0306951999664307,
|
| 1072 |
+
"logps/rejected": -5.601127624511719,
|
| 1073 |
+
"loss": 14.5714,
|
| 1074 |
+
"nll_loss": 1.1835569143295288,
|
| 1075 |
+
"po_loss": -0.7282010316848755,
|
| 1076 |
+
"rewards/accuracies": 0.925000011920929,
|
| 1077 |
+
"rewards/chosen": -0.5153475999832153,
|
| 1078 |
+
"rewards/margins": 2.2852160930633545,
|
| 1079 |
+
"rewards/rejected": -2.8005638122558594,
|
| 1080 |
+
"step": 315
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 2.7168,
|
| 1084 |
+
"grad_norm": 337.47796630859375,
|
| 1085 |
+
"learning_rate": 1.1853588439213442e-06,
|
| 1086 |
+
"log_odds_chosen": 6.4345574378967285,
|
| 1087 |
+
"log_odds_ratio": -0.23376791179180145,
|
| 1088 |
+
"logps/chosen": -1.014891266822815,
|
| 1089 |
+
"logps/rejected": -7.0141282081604,
|
| 1090 |
+
"loss": 16.8373,
|
| 1091 |
+
"nll_loss": 1.2330230474472046,
|
| 1092 |
+
"po_loss": -0.7068589329719543,
|
| 1093 |
+
"rewards/accuracies": 0.9437500238418579,
|
| 1094 |
+
"rewards/chosen": -0.5074456334114075,
|
| 1095 |
+
"rewards/margins": 2.9996187686920166,
|
| 1096 |
+
"rewards/rejected": -3.5070641040802,
|
| 1097 |
+
"step": 320
|
| 1098 |
+
},
|
| 1099 |
+
{
|
| 1100 |
+
"epoch": 2.7594666666666665,
|
| 1101 |
+
"grad_norm": 251.52114868164062,
|
| 1102 |
+
"learning_rate": 8.357964040363209e-07,
|
| 1103 |
+
"log_odds_chosen": 7.006105899810791,
|
| 1104 |
+
"log_odds_ratio": -0.3340076804161072,
|
| 1105 |
+
"logps/chosen": -1.0801838636398315,
|
| 1106 |
+
"logps/rejected": -7.584368705749512,
|
| 1107 |
+
"loss": 7.9524,
|
| 1108 |
+
"nll_loss": 1.2427351474761963,
|
| 1109 |
+
"po_loss": -0.9942227602005005,
|
| 1110 |
+
"rewards/accuracies": 0.9281250238418579,
|
| 1111 |
+
"rewards/chosen": -0.5400919318199158,
|
| 1112 |
+
"rewards/margins": 3.2520930767059326,
|
| 1113 |
+
"rewards/rejected": -3.792184352874756,
|
| 1114 |
+
"step": 325
|
| 1115 |
+
},
|
| 1116 |
+
{
|
| 1117 |
+
"epoch": 2.8021333333333334,
|
| 1118 |
+
"grad_norm": 500.19580078125,
|
| 1119 |
+
"learning_rate": 5.463099816548579e-07,
|
| 1120 |
+
"log_odds_chosen": 7.333872318267822,
|
| 1121 |
+
"log_odds_ratio": -0.251155823469162,
|
| 1122 |
+
"logps/chosen": -1.1577497720718384,
|
| 1123 |
+
"logps/rejected": -8.027981758117676,
|
| 1124 |
+
"loss": 19.0267,
|
| 1125 |
+
"nll_loss": 1.2874268293380737,
|
| 1126 |
+
"po_loss": -0.6928423047065735,
|
| 1127 |
+
"rewards/accuracies": 0.9312499761581421,
|
| 1128 |
+
"rewards/chosen": -0.5788748860359192,
|
| 1129 |
+
"rewards/margins": 3.4351158142089844,
|
| 1130 |
+
"rewards/rejected": -4.013990879058838,
|
| 1131 |
+
"step": 330
|
| 1132 |
+
},
|
| 1133 |
+
{
|
| 1134 |
+
"epoch": 2.8448,
|
| 1135 |
+
"grad_norm": 827.9666748046875,
|
| 1136 |
+
"learning_rate": 3.1761928563510955e-07,
|
| 1137 |
+
"log_odds_chosen": 5.787890434265137,
|
| 1138 |
+
"log_odds_ratio": -0.2869322896003723,
|
| 1139 |
+
"logps/chosen": -1.0971204042434692,
|
| 1140 |
+
"logps/rejected": -6.4105658531188965,
|
| 1141 |
+
"loss": 12.9881,
|
| 1142 |
+
"nll_loss": 1.2217962741851807,
|
| 1143 |
+
"po_loss": -0.8159177899360657,
|
| 1144 |
+
"rewards/accuracies": 0.9125000238418579,
|
| 1145 |
+
"rewards/chosen": -0.5485602021217346,
|
| 1146 |
+
"rewards/margins": 2.6567225456237793,
|
| 1147 |
+
"rewards/rejected": -3.2052829265594482,
|
| 1148 |
+
"step": 335
|
| 1149 |
+
},
|
| 1150 |
+
{
|
| 1151 |
+
"epoch": 2.8874666666666666,
|
| 1152 |
+
"grad_norm": 292.38897705078125,
|
| 1153 |
+
"learning_rate": 1.5029287708036854e-07,
|
| 1154 |
+
"log_odds_chosen": 6.05283260345459,
|
| 1155 |
+
"log_odds_ratio": -0.24598774313926697,
|
| 1156 |
+
"logps/chosen": -1.0574235916137695,
|
| 1157 |
+
"logps/rejected": -6.646490573883057,
|
| 1158 |
+
"loss": 7.9611,
|
| 1159 |
+
"nll_loss": 1.1931092739105225,
|
| 1160 |
+
"po_loss": -0.9443262219429016,
|
| 1161 |
+
"rewards/accuracies": 0.956250011920929,
|
| 1162 |
+
"rewards/chosen": -0.5287117958068848,
|
| 1163 |
+
"rewards/margins": 2.7945332527160645,
|
| 1164 |
+
"rewards/rejected": -3.3232452869415283,
|
| 1165 |
+
"step": 340
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 2.9301333333333335,
|
| 1169 |
+
"grad_norm": 387.1372985839844,
|
| 1170 |
+
"learning_rate": 4.474675580662113e-08,
|
| 1171 |
+
"log_odds_chosen": 6.334747791290283,
|
| 1172 |
+
"log_odds_ratio": -0.32026782631874084,
|
| 1173 |
+
"logps/chosen": -1.1003960371017456,
|
| 1174 |
+
"logps/rejected": -6.976574897766113,
|
| 1175 |
+
"loss": 12.2009,
|
| 1176 |
+
"nll_loss": 1.235642671585083,
|
| 1177 |
+
"po_loss": -0.8543656468391418,
|
| 1178 |
+
"rewards/accuracies": 0.9156249761581421,
|
| 1179 |
+
"rewards/chosen": -0.5501980185508728,
|
| 1180 |
+
"rewards/margins": 2.938089370727539,
|
| 1181 |
+
"rewards/rejected": -3.4882874488830566,
|
| 1182 |
+
"step": 345
|
| 1183 |
+
},
|
| 1184 |
+
{
|
| 1185 |
+
"epoch": 2.9728,
|
| 1186 |
+
"grad_norm": 1250.0882568359375,
|
| 1187 |
+
"learning_rate": 1.2433261014244136e-09,
|
| 1188 |
+
"log_odds_chosen": 7.907639980316162,
|
| 1189 |
+
"log_odds_ratio": -0.22040143609046936,
|
| 1190 |
+
"logps/chosen": -1.2740892171859741,
|
| 1191 |
+
"logps/rejected": -8.710322380065918,
|
| 1192 |
+
"loss": 9.5303,
|
| 1193 |
+
"nll_loss": 1.3264214992523193,
|
| 1194 |
+
"po_loss": -1.028598666191101,
|
| 1195 |
+
"rewards/accuracies": 0.9281250238418579,
|
| 1196 |
+
"rewards/chosen": -0.6370446085929871,
|
| 1197 |
+
"rewards/margins": 3.7181167602539062,
|
| 1198 |
+
"rewards/rejected": -4.355161190032959,
|
| 1199 |
+
"step": 350
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 2.981333333333333,
|
| 1203 |
+
"step": 351,
|
| 1204 |
+
"total_flos": 0.0,
|
| 1205 |
+
"train_loss": 54.80183921716152,
|
| 1206 |
+
"train_runtime": 10133.5503,
|
| 1207 |
+
"train_samples_per_second": 2.22,
|
| 1208 |
+
"train_steps_per_second": 0.035
|
| 1209 |
+
}
|
| 1210 |
+
],
|
| 1211 |
+
"logging_steps": 5,
|
| 1212 |
+
"max_steps": 351,
|
| 1213 |
+
"num_input_tokens_seen": 0,
|
| 1214 |
+
"num_train_epochs": 3,
|
| 1215 |
+
"save_steps": 100000,
|
| 1216 |
+
"stateful_callbacks": {
|
| 1217 |
+
"TrainerControl": {
|
| 1218 |
+
"args": {
|
| 1219 |
+
"should_epoch_stop": false,
|
| 1220 |
+
"should_evaluate": false,
|
| 1221 |
+
"should_log": false,
|
| 1222 |
+
"should_save": true,
|
| 1223 |
+
"should_training_stop": true
|
| 1224 |
+
},
|
| 1225 |
+
"attributes": {}
|
| 1226 |
+
}
|
| 1227 |
+
},
|
| 1228 |
+
"total_flos": 0.0,
|
| 1229 |
+
"train_batch_size": 1,
|
| 1230 |
+
"trial_name": null,
|
| 1231 |
+
"trial_params": null
|
| 1232 |
+
}
|