Model save
Browse files- README.md +68 -0
- all_results.json +9 -0
- generation_config.json +6 -0
- train_results.json +9 -0
- trainer_state.json +2910 -0
README.md
ADDED
|
@@ -0,0 +1,68 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
base_model: AmberYifan/Qwen2-7B-sft-ultrachat
|
| 3 |
+
library_name: transformers
|
| 4 |
+
model_name: Qwen2-7B-sft-ultrachat-peers-pool
|
| 5 |
+
tags:
|
| 6 |
+
- generated_from_trainer
|
| 7 |
+
- trl
|
| 8 |
+
- dpo
|
| 9 |
+
licence: license
|
| 10 |
+
---
|
| 11 |
+
|
| 12 |
+
# Model Card for Qwen2-7B-sft-ultrachat-peers-pool
|
| 13 |
+
|
| 14 |
+
This model is a fine-tuned version of [AmberYifan/Qwen2-7B-sft-ultrachat](https://huggingface.co/AmberYifan/Qwen2-7B-sft-ultrachat).
|
| 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="AmberYifan/Qwen2-7B-sft-ultrachat-peers-pool", 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/yifanwang/huggingface/runs/tgj30sw5)
|
| 31 |
+
|
| 32 |
+
This model was trained with DPO, a method introduced in [Direct Preference Optimization: Your Language Model is Secretly a Reward Model](https://huggingface.co/papers/2305.18290).
|
| 33 |
+
|
| 34 |
+
### Framework versions
|
| 35 |
+
|
| 36 |
+
- TRL: 0.12.2
|
| 37 |
+
- Transformers: 4.46.3
|
| 38 |
+
- Pytorch: 2.6.0
|
| 39 |
+
- Datasets: 3.3.0
|
| 40 |
+
- Tokenizers: 0.20.3
|
| 41 |
+
|
| 42 |
+
## Citations
|
| 43 |
+
|
| 44 |
+
Cite DPO as:
|
| 45 |
+
|
| 46 |
+
```bibtex
|
| 47 |
+
@inproceedings{rafailov2023direct,
|
| 48 |
+
title = {{Direct Preference Optimization: Your Language Model is Secretly a Reward Model}},
|
| 49 |
+
author = {Rafael Rafailov and Archit Sharma and Eric Mitchell and Christopher D. Manning and Stefano Ermon and Chelsea Finn},
|
| 50 |
+
year = 2023,
|
| 51 |
+
booktitle = {Advances in Neural Information Processing Systems 36: Annual Conference on Neural Information Processing Systems 2023, NeurIPS 2023, New Orleans, LA, USA, December 10 - 16, 2023},
|
| 52 |
+
url = {http://papers.nips.cc/paper_files/paper/2023/hash/a85b405ed65c6477a4fe8302b5e06ce7-Abstract-Conference.html},
|
| 53 |
+
editor = {Alice Oh and Tristan Naumann and Amir Globerson and Kate Saenko and Moritz Hardt and Sergey Levine},
|
| 54 |
+
}
|
| 55 |
+
```
|
| 56 |
+
|
| 57 |
+
Cite TRL as:
|
| 58 |
+
|
| 59 |
+
```bibtex
|
| 60 |
+
@misc{vonwerra2022trl,
|
| 61 |
+
title = {{TRL: Transformer Reinforcement Learning}},
|
| 62 |
+
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},
|
| 63 |
+
year = 2020,
|
| 64 |
+
journal = {GitHub repository},
|
| 65 |
+
publisher = {GitHub},
|
| 66 |
+
howpublished = {\url{https://github.com/huggingface/trl}}
|
| 67 |
+
}
|
| 68 |
+
```
|
all_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"total_flos": 0.0,
|
| 4 |
+
"train_loss": 0.14497800362415802,
|
| 5 |
+
"train_runtime": 10901.8033,
|
| 6 |
+
"train_samples": 9977,
|
| 7 |
+
"train_samples_per_second": 2.746,
|
| 8 |
+
"train_steps_per_second": 0.172
|
| 9 |
+
}
|
generation_config.json
ADDED
|
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token_id": 151643,
|
| 3 |
+
"eos_token_id": 151643,
|
| 4 |
+
"max_new_tokens": 2048,
|
| 5 |
+
"transformers_version": "4.46.3"
|
| 6 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"total_flos": 0.0,
|
| 4 |
+
"train_loss": 0.14497800362415802,
|
| 5 |
+
"train_runtime": 10901.8033,
|
| 6 |
+
"train_samples": 9977,
|
| 7 |
+
"train_samples_per_second": 2.746,
|
| 8 |
+
"train_steps_per_second": 0.172
|
| 9 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2910 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 3.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 1872,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.0016025641025641025,
|
| 13 |
+
"grad_norm": 81.23951893705919,
|
| 14 |
+
"learning_rate": 2.6595744680851063e-09,
|
| 15 |
+
"logits/chosen": 0.1357421875,
|
| 16 |
+
"logits/rejected": -0.2060546875,
|
| 17 |
+
"logps/chosen": -336.0,
|
| 18 |
+
"logps/rejected": -346.0,
|
| 19 |
+
"loss": 0.6914,
|
| 20 |
+
"rewards/accuracies": 0.0,
|
| 21 |
+
"rewards/chosen": 0.0,
|
| 22 |
+
"rewards/margins": 0.0,
|
| 23 |
+
"rewards/rejected": 0.0,
|
| 24 |
+
"step": 1
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.016025641025641024,
|
| 28 |
+
"grad_norm": 86.32419393322078,
|
| 29 |
+
"learning_rate": 2.6595744680851062e-08,
|
| 30 |
+
"logits/chosen": -0.16015625,
|
| 31 |
+
"logits/rejected": -0.474609375,
|
| 32 |
+
"logps/chosen": -364.0,
|
| 33 |
+
"logps/rejected": -312.0,
|
| 34 |
+
"loss": 0.702,
|
| 35 |
+
"rewards/accuracies": 0.1527777761220932,
|
| 36 |
+
"rewards/chosen": -0.009033203125,
|
| 37 |
+
"rewards/margins": -0.01806640625,
|
| 38 |
+
"rewards/rejected": 0.009033203125,
|
| 39 |
+
"step": 10
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 0.03205128205128205,
|
| 43 |
+
"grad_norm": 100.77964496481692,
|
| 44 |
+
"learning_rate": 5.3191489361702123e-08,
|
| 45 |
+
"logits/chosen": -0.16796875,
|
| 46 |
+
"logits/rejected": -0.466796875,
|
| 47 |
+
"logps/chosen": -376.0,
|
| 48 |
+
"logps/rejected": -310.0,
|
| 49 |
+
"loss": 0.6939,
|
| 50 |
+
"rewards/accuracies": 0.21250000596046448,
|
| 51 |
+
"rewards/chosen": -0.006561279296875,
|
| 52 |
+
"rewards/margins": -0.0021820068359375,
|
| 53 |
+
"rewards/rejected": -0.00439453125,
|
| 54 |
+
"step": 20
|
| 55 |
+
},
|
| 56 |
+
{
|
| 57 |
+
"epoch": 0.04807692307692308,
|
| 58 |
+
"grad_norm": 98.51227524666062,
|
| 59 |
+
"learning_rate": 7.978723404255319e-08,
|
| 60 |
+
"logits/chosen": -0.22265625,
|
| 61 |
+
"logits/rejected": -0.5234375,
|
| 62 |
+
"logps/chosen": -390.0,
|
| 63 |
+
"logps/rejected": -332.0,
|
| 64 |
+
"loss": 0.6834,
|
| 65 |
+
"rewards/accuracies": 0.3375000059604645,
|
| 66 |
+
"rewards/chosen": 0.0419921875,
|
| 67 |
+
"rewards/margins": 0.03515625,
|
| 68 |
+
"rewards/rejected": 0.00689697265625,
|
| 69 |
+
"step": 30
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 0.0641025641025641,
|
| 73 |
+
"grad_norm": 73.06116455998657,
|
| 74 |
+
"learning_rate": 1.0638297872340425e-07,
|
| 75 |
+
"logits/chosen": -0.23828125,
|
| 76 |
+
"logits/rejected": -0.4296875,
|
| 77 |
+
"logps/chosen": -370.0,
|
| 78 |
+
"logps/rejected": -304.0,
|
| 79 |
+
"loss": 0.6555,
|
| 80 |
+
"rewards/accuracies": 0.5375000238418579,
|
| 81 |
+
"rewards/chosen": 0.142578125,
|
| 82 |
+
"rewards/margins": 0.09521484375,
|
| 83 |
+
"rewards/rejected": 0.046875,
|
| 84 |
+
"step": 40
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 0.08012820512820513,
|
| 88 |
+
"grad_norm": 76.17217165247452,
|
| 89 |
+
"learning_rate": 1.329787234042553e-07,
|
| 90 |
+
"logits/chosen": -0.2421875,
|
| 91 |
+
"logits/rejected": -0.46875,
|
| 92 |
+
"logps/chosen": -382.0,
|
| 93 |
+
"logps/rejected": -302.0,
|
| 94 |
+
"loss": 0.6393,
|
| 95 |
+
"rewards/accuracies": 0.6875,
|
| 96 |
+
"rewards/chosen": 0.2578125,
|
| 97 |
+
"rewards/margins": 0.177734375,
|
| 98 |
+
"rewards/rejected": 0.07958984375,
|
| 99 |
+
"step": 50
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"epoch": 0.09615384615384616,
|
| 103 |
+
"grad_norm": 61.759030001028826,
|
| 104 |
+
"learning_rate": 1.5957446808510638e-07,
|
| 105 |
+
"logits/chosen": -0.0164794921875,
|
| 106 |
+
"logits/rejected": -0.291015625,
|
| 107 |
+
"logps/chosen": -358.0,
|
| 108 |
+
"logps/rejected": -306.0,
|
| 109 |
+
"loss": 0.6168,
|
| 110 |
+
"rewards/accuracies": 0.675000011920929,
|
| 111 |
+
"rewards/chosen": 0.376953125,
|
| 112 |
+
"rewards/margins": 0.2255859375,
|
| 113 |
+
"rewards/rejected": 0.15234375,
|
| 114 |
+
"step": 60
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.11217948717948718,
|
| 118 |
+
"grad_norm": 64.71686861478489,
|
| 119 |
+
"learning_rate": 1.8617021276595742e-07,
|
| 120 |
+
"logits/chosen": -0.283203125,
|
| 121 |
+
"logits/rejected": -0.50390625,
|
| 122 |
+
"logps/chosen": -396.0,
|
| 123 |
+
"logps/rejected": -292.0,
|
| 124 |
+
"loss": 0.5847,
|
| 125 |
+
"rewards/accuracies": 0.800000011920929,
|
| 126 |
+
"rewards/chosen": 0.58203125,
|
| 127 |
+
"rewards/margins": 0.37890625,
|
| 128 |
+
"rewards/rejected": 0.201171875,
|
| 129 |
+
"step": 70
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.1282051282051282,
|
| 133 |
+
"grad_norm": 52.52316366702727,
|
| 134 |
+
"learning_rate": 2.127659574468085e-07,
|
| 135 |
+
"logits/chosen": -0.35546875,
|
| 136 |
+
"logits/rejected": -0.66796875,
|
| 137 |
+
"logps/chosen": -368.0,
|
| 138 |
+
"logps/rejected": -302.0,
|
| 139 |
+
"loss": 0.5555,
|
| 140 |
+
"rewards/accuracies": 0.737500011920929,
|
| 141 |
+
"rewards/chosen": 0.6640625,
|
| 142 |
+
"rewards/margins": 0.484375,
|
| 143 |
+
"rewards/rejected": 0.1787109375,
|
| 144 |
+
"step": 80
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 0.14423076923076922,
|
| 148 |
+
"grad_norm": 59.670100120011845,
|
| 149 |
+
"learning_rate": 2.393617021276596e-07,
|
| 150 |
+
"logits/chosen": -0.251953125,
|
| 151 |
+
"logits/rejected": -0.55078125,
|
| 152 |
+
"logps/chosen": -388.0,
|
| 153 |
+
"logps/rejected": -312.0,
|
| 154 |
+
"loss": 0.539,
|
| 155 |
+
"rewards/accuracies": 0.7749999761581421,
|
| 156 |
+
"rewards/chosen": 0.8046875,
|
| 157 |
+
"rewards/margins": 0.5703125,
|
| 158 |
+
"rewards/rejected": 0.2353515625,
|
| 159 |
+
"step": 90
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 0.16025641025641027,
|
| 163 |
+
"grad_norm": 66.87233629909264,
|
| 164 |
+
"learning_rate": 2.659574468085106e-07,
|
| 165 |
+
"logits/chosen": -0.314453125,
|
| 166 |
+
"logits/rejected": -0.478515625,
|
| 167 |
+
"logps/chosen": -346.0,
|
| 168 |
+
"logps/rejected": -324.0,
|
| 169 |
+
"loss": 0.4916,
|
| 170 |
+
"rewards/accuracies": 0.699999988079071,
|
| 171 |
+
"rewards/chosen": 0.85546875,
|
| 172 |
+
"rewards/margins": 0.48046875,
|
| 173 |
+
"rewards/rejected": 0.373046875,
|
| 174 |
+
"step": 100
|
| 175 |
+
},
|
| 176 |
+
{
|
| 177 |
+
"epoch": 0.1762820512820513,
|
| 178 |
+
"grad_norm": 53.961384431262026,
|
| 179 |
+
"learning_rate": 2.925531914893617e-07,
|
| 180 |
+
"logits/chosen": -0.294921875,
|
| 181 |
+
"logits/rejected": -0.65625,
|
| 182 |
+
"logps/chosen": -386.0,
|
| 183 |
+
"logps/rejected": -306.0,
|
| 184 |
+
"loss": 0.5006,
|
| 185 |
+
"rewards/accuracies": 0.824999988079071,
|
| 186 |
+
"rewards/chosen": 0.9140625,
|
| 187 |
+
"rewards/margins": 0.640625,
|
| 188 |
+
"rewards/rejected": 0.275390625,
|
| 189 |
+
"step": 110
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 0.19230769230769232,
|
| 193 |
+
"grad_norm": 49.6732649258676,
|
| 194 |
+
"learning_rate": 3.1914893617021275e-07,
|
| 195 |
+
"logits/chosen": -0.298828125,
|
| 196 |
+
"logits/rejected": -0.40234375,
|
| 197 |
+
"logps/chosen": -388.0,
|
| 198 |
+
"logps/rejected": -308.0,
|
| 199 |
+
"loss": 0.4499,
|
| 200 |
+
"rewards/accuracies": 0.7250000238418579,
|
| 201 |
+
"rewards/chosen": 1.1171875,
|
| 202 |
+
"rewards/margins": 0.82421875,
|
| 203 |
+
"rewards/rejected": 0.29296875,
|
| 204 |
+
"step": 120
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 0.20833333333333334,
|
| 208 |
+
"grad_norm": 52.819440458631206,
|
| 209 |
+
"learning_rate": 3.457446808510638e-07,
|
| 210 |
+
"logits/chosen": -0.1591796875,
|
| 211 |
+
"logits/rejected": -0.28515625,
|
| 212 |
+
"logps/chosen": -374.0,
|
| 213 |
+
"logps/rejected": -280.0,
|
| 214 |
+
"loss": 0.4087,
|
| 215 |
+
"rewards/accuracies": 0.887499988079071,
|
| 216 |
+
"rewards/chosen": 1.3671875,
|
| 217 |
+
"rewards/margins": 1.203125,
|
| 218 |
+
"rewards/rejected": 0.16015625,
|
| 219 |
+
"step": 130
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.22435897435897437,
|
| 223 |
+
"grad_norm": 45.055792147900604,
|
| 224 |
+
"learning_rate": 3.7234042553191484e-07,
|
| 225 |
+
"logits/chosen": -0.2177734375,
|
| 226 |
+
"logits/rejected": -0.380859375,
|
| 227 |
+
"logps/chosen": -364.0,
|
| 228 |
+
"logps/rejected": -300.0,
|
| 229 |
+
"loss": 0.4263,
|
| 230 |
+
"rewards/accuracies": 0.862500011920929,
|
| 231 |
+
"rewards/chosen": 1.328125,
|
| 232 |
+
"rewards/margins": 1.1640625,
|
| 233 |
+
"rewards/rejected": 0.1650390625,
|
| 234 |
+
"step": 140
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.2403846153846154,
|
| 238 |
+
"grad_norm": 37.20442727443192,
|
| 239 |
+
"learning_rate": 3.989361702127659e-07,
|
| 240 |
+
"logits/chosen": -0.208984375,
|
| 241 |
+
"logits/rejected": -0.421875,
|
| 242 |
+
"logps/chosen": -392.0,
|
| 243 |
+
"logps/rejected": -310.0,
|
| 244 |
+
"loss": 0.4273,
|
| 245 |
+
"rewards/accuracies": 0.8500000238418579,
|
| 246 |
+
"rewards/chosen": 1.3671875,
|
| 247 |
+
"rewards/margins": 1.265625,
|
| 248 |
+
"rewards/rejected": 0.10205078125,
|
| 249 |
+
"step": 150
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 0.2564102564102564,
|
| 253 |
+
"grad_norm": 67.23258640755056,
|
| 254 |
+
"learning_rate": 4.25531914893617e-07,
|
| 255 |
+
"logits/chosen": -0.1416015625,
|
| 256 |
+
"logits/rejected": -0.365234375,
|
| 257 |
+
"logps/chosen": -380.0,
|
| 258 |
+
"logps/rejected": -322.0,
|
| 259 |
+
"loss": 0.3995,
|
| 260 |
+
"rewards/accuracies": 0.7875000238418579,
|
| 261 |
+
"rewards/chosen": 1.1484375,
|
| 262 |
+
"rewards/margins": 1.1953125,
|
| 263 |
+
"rewards/rejected": -0.047119140625,
|
| 264 |
+
"step": 160
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 0.2724358974358974,
|
| 268 |
+
"grad_norm": 76.82700510564892,
|
| 269 |
+
"learning_rate": 4.5212765957446806e-07,
|
| 270 |
+
"logits/chosen": -0.1376953125,
|
| 271 |
+
"logits/rejected": -0.42578125,
|
| 272 |
+
"logps/chosen": -360.0,
|
| 273 |
+
"logps/rejected": -312.0,
|
| 274 |
+
"loss": 0.4255,
|
| 275 |
+
"rewards/accuracies": 0.800000011920929,
|
| 276 |
+
"rewards/chosen": 1.1171875,
|
| 277 |
+
"rewards/margins": 1.125,
|
| 278 |
+
"rewards/rejected": -0.0093994140625,
|
| 279 |
+
"step": 170
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 0.28846153846153844,
|
| 283 |
+
"grad_norm": 57.73247757369804,
|
| 284 |
+
"learning_rate": 4.787234042553192e-07,
|
| 285 |
+
"logits/chosen": -0.083984375,
|
| 286 |
+
"logits/rejected": -0.1953125,
|
| 287 |
+
"logps/chosen": -368.0,
|
| 288 |
+
"logps/rejected": -316.0,
|
| 289 |
+
"loss": 0.3897,
|
| 290 |
+
"rewards/accuracies": 0.800000011920929,
|
| 291 |
+
"rewards/chosen": 1.3125,
|
| 292 |
+
"rewards/margins": 1.328125,
|
| 293 |
+
"rewards/rejected": -0.01129150390625,
|
| 294 |
+
"step": 180
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 0.30448717948717946,
|
| 298 |
+
"grad_norm": 93.3245464788157,
|
| 299 |
+
"learning_rate": 4.994061757719714e-07,
|
| 300 |
+
"logits/chosen": -0.158203125,
|
| 301 |
+
"logits/rejected": -0.408203125,
|
| 302 |
+
"logps/chosen": -344.0,
|
| 303 |
+
"logps/rejected": -318.0,
|
| 304 |
+
"loss": 0.4427,
|
| 305 |
+
"rewards/accuracies": 0.8500000238418579,
|
| 306 |
+
"rewards/chosen": 1.3828125,
|
| 307 |
+
"rewards/margins": 1.515625,
|
| 308 |
+
"rewards/rejected": -0.130859375,
|
| 309 |
+
"step": 190
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 0.32051282051282054,
|
| 313 |
+
"grad_norm": 31.0057130908168,
|
| 314 |
+
"learning_rate": 4.96437054631829e-07,
|
| 315 |
+
"logits/chosen": -0.185546875,
|
| 316 |
+
"logits/rejected": -0.333984375,
|
| 317 |
+
"logps/chosen": -398.0,
|
| 318 |
+
"logps/rejected": -276.0,
|
| 319 |
+
"loss": 0.3853,
|
| 320 |
+
"rewards/accuracies": 0.8999999761581421,
|
| 321 |
+
"rewards/chosen": 1.2578125,
|
| 322 |
+
"rewards/margins": 1.5546875,
|
| 323 |
+
"rewards/rejected": -0.298828125,
|
| 324 |
+
"step": 200
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.33653846153846156,
|
| 328 |
+
"grad_norm": 31.866164526406457,
|
| 329 |
+
"learning_rate": 4.934679334916864e-07,
|
| 330 |
+
"logits/chosen": -0.134765625,
|
| 331 |
+
"logits/rejected": -0.427734375,
|
| 332 |
+
"logps/chosen": -372.0,
|
| 333 |
+
"logps/rejected": -334.0,
|
| 334 |
+
"loss": 0.2608,
|
| 335 |
+
"rewards/accuracies": 0.9125000238418579,
|
| 336 |
+
"rewards/chosen": 1.4375,
|
| 337 |
+
"rewards/margins": 1.703125,
|
| 338 |
+
"rewards/rejected": -0.267578125,
|
| 339 |
+
"step": 210
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.3525641025641026,
|
| 343 |
+
"grad_norm": 26.376971522466825,
|
| 344 |
+
"learning_rate": 4.904988123515439e-07,
|
| 345 |
+
"logits/chosen": -0.0673828125,
|
| 346 |
+
"logits/rejected": -0.150390625,
|
| 347 |
+
"logps/chosen": -392.0,
|
| 348 |
+
"logps/rejected": -294.0,
|
| 349 |
+
"loss": 0.313,
|
| 350 |
+
"rewards/accuracies": 0.875,
|
| 351 |
+
"rewards/chosen": 1.6796875,
|
| 352 |
+
"rewards/margins": 1.921875,
|
| 353 |
+
"rewards/rejected": -0.2451171875,
|
| 354 |
+
"step": 220
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 0.3685897435897436,
|
| 358 |
+
"grad_norm": 19.856366106787778,
|
| 359 |
+
"learning_rate": 4.875296912114014e-07,
|
| 360 |
+
"logits/chosen": -0.068359375,
|
| 361 |
+
"logits/rejected": -0.373046875,
|
| 362 |
+
"logps/chosen": -342.0,
|
| 363 |
+
"logps/rejected": -298.0,
|
| 364 |
+
"loss": 0.2838,
|
| 365 |
+
"rewards/accuracies": 0.875,
|
| 366 |
+
"rewards/chosen": 1.46875,
|
| 367 |
+
"rewards/margins": 2.125,
|
| 368 |
+
"rewards/rejected": -0.65625,
|
| 369 |
+
"step": 230
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 0.38461538461538464,
|
| 373 |
+
"grad_norm": 55.84167138753407,
|
| 374 |
+
"learning_rate": 4.845605700712589e-07,
|
| 375 |
+
"logits/chosen": -0.12890625,
|
| 376 |
+
"logits/rejected": -0.443359375,
|
| 377 |
+
"logps/chosen": -368.0,
|
| 378 |
+
"logps/rejected": -322.0,
|
| 379 |
+
"loss": 0.2748,
|
| 380 |
+
"rewards/accuracies": 0.862500011920929,
|
| 381 |
+
"rewards/chosen": 1.5546875,
|
| 382 |
+
"rewards/margins": 2.5,
|
| 383 |
+
"rewards/rejected": -0.953125,
|
| 384 |
+
"step": 240
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 0.40064102564102566,
|
| 388 |
+
"grad_norm": 37.23188688423797,
|
| 389 |
+
"learning_rate": 4.815914489311164e-07,
|
| 390 |
+
"logits/chosen": -0.1357421875,
|
| 391 |
+
"logits/rejected": -0.427734375,
|
| 392 |
+
"logps/chosen": -364.0,
|
| 393 |
+
"logps/rejected": -324.0,
|
| 394 |
+
"loss": 0.2543,
|
| 395 |
+
"rewards/accuracies": 0.9375,
|
| 396 |
+
"rewards/chosen": 1.6640625,
|
| 397 |
+
"rewards/margins": 2.734375,
|
| 398 |
+
"rewards/rejected": -1.0703125,
|
| 399 |
+
"step": 250
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 0.4166666666666667,
|
| 403 |
+
"grad_norm": 35.660892697083206,
|
| 404 |
+
"learning_rate": 4.786223277909738e-07,
|
| 405 |
+
"logits/chosen": -0.123046875,
|
| 406 |
+
"logits/rejected": -0.470703125,
|
| 407 |
+
"logps/chosen": -366.0,
|
| 408 |
+
"logps/rejected": -310.0,
|
| 409 |
+
"loss": 0.2583,
|
| 410 |
+
"rewards/accuracies": 0.862500011920929,
|
| 411 |
+
"rewards/chosen": 1.1484375,
|
| 412 |
+
"rewards/margins": 2.375,
|
| 413 |
+
"rewards/rejected": -1.234375,
|
| 414 |
+
"step": 260
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 0.4326923076923077,
|
| 418 |
+
"grad_norm": 46.54757231137472,
|
| 419 |
+
"learning_rate": 4.756532066508313e-07,
|
| 420 |
+
"logits/chosen": -0.2021484375,
|
| 421 |
+
"logits/rejected": -0.328125,
|
| 422 |
+
"logps/chosen": -374.0,
|
| 423 |
+
"logps/rejected": -340.0,
|
| 424 |
+
"loss": 0.4125,
|
| 425 |
+
"rewards/accuracies": 0.762499988079071,
|
| 426 |
+
"rewards/chosen": 1.1015625,
|
| 427 |
+
"rewards/margins": 1.9765625,
|
| 428 |
+
"rewards/rejected": -0.875,
|
| 429 |
+
"step": 270
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.44871794871794873,
|
| 433 |
+
"grad_norm": 37.544511359682254,
|
| 434 |
+
"learning_rate": 4.7268408551068883e-07,
|
| 435 |
+
"logits/chosen": 0.03857421875,
|
| 436 |
+
"logits/rejected": -0.54296875,
|
| 437 |
+
"logps/chosen": -384.0,
|
| 438 |
+
"logps/rejected": -342.0,
|
| 439 |
+
"loss": 0.2471,
|
| 440 |
+
"rewards/accuracies": 0.862500011920929,
|
| 441 |
+
"rewards/chosen": 1.1796875,
|
| 442 |
+
"rewards/margins": 2.5,
|
| 443 |
+
"rewards/rejected": -1.3203125,
|
| 444 |
+
"step": 280
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 0.46474358974358976,
|
| 448 |
+
"grad_norm": 56.789270954370565,
|
| 449 |
+
"learning_rate": 4.697149643705463e-07,
|
| 450 |
+
"logits/chosen": -0.07958984375,
|
| 451 |
+
"logits/rejected": -0.29296875,
|
| 452 |
+
"logps/chosen": -348.0,
|
| 453 |
+
"logps/rejected": -328.0,
|
| 454 |
+
"loss": 0.297,
|
| 455 |
+
"rewards/accuracies": 0.887499988079071,
|
| 456 |
+
"rewards/chosen": 1.4140625,
|
| 457 |
+
"rewards/margins": 2.46875,
|
| 458 |
+
"rewards/rejected": -1.0546875,
|
| 459 |
+
"step": 290
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 0.4807692307692308,
|
| 463 |
+
"grad_norm": 14.22992572113077,
|
| 464 |
+
"learning_rate": 4.667458432304038e-07,
|
| 465 |
+
"logits/chosen": -0.2021484375,
|
| 466 |
+
"logits/rejected": -0.388671875,
|
| 467 |
+
"logps/chosen": -372.0,
|
| 468 |
+
"logps/rejected": -322.0,
|
| 469 |
+
"loss": 0.2564,
|
| 470 |
+
"rewards/accuracies": 0.887499988079071,
|
| 471 |
+
"rewards/chosen": 1.6796875,
|
| 472 |
+
"rewards/margins": 2.765625,
|
| 473 |
+
"rewards/rejected": -1.0859375,
|
| 474 |
+
"step": 300
|
| 475 |
+
},
|
| 476 |
+
{
|
| 477 |
+
"epoch": 0.4967948717948718,
|
| 478 |
+
"grad_norm": 36.87971897536812,
|
| 479 |
+
"learning_rate": 4.6377672209026127e-07,
|
| 480 |
+
"logits/chosen": -0.10888671875,
|
| 481 |
+
"logits/rejected": -0.412109375,
|
| 482 |
+
"logps/chosen": -390.0,
|
| 483 |
+
"logps/rejected": -330.0,
|
| 484 |
+
"loss": 0.1529,
|
| 485 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 486 |
+
"rewards/chosen": 2.109375,
|
| 487 |
+
"rewards/margins": 3.328125,
|
| 488 |
+
"rewards/rejected": -1.2109375,
|
| 489 |
+
"step": 310
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 0.5128205128205128,
|
| 493 |
+
"grad_norm": 31.81729836940978,
|
| 494 |
+
"learning_rate": 4.6080760095011875e-07,
|
| 495 |
+
"logits/chosen": -0.154296875,
|
| 496 |
+
"logits/rejected": -0.35546875,
|
| 497 |
+
"logps/chosen": -374.0,
|
| 498 |
+
"logps/rejected": -304.0,
|
| 499 |
+
"loss": 0.26,
|
| 500 |
+
"rewards/accuracies": 0.887499988079071,
|
| 501 |
+
"rewards/chosen": 1.859375,
|
| 502 |
+
"rewards/margins": 2.875,
|
| 503 |
+
"rewards/rejected": -1.0234375,
|
| 504 |
+
"step": 320
|
| 505 |
+
},
|
| 506 |
+
{
|
| 507 |
+
"epoch": 0.5288461538461539,
|
| 508 |
+
"grad_norm": 27.609497368983103,
|
| 509 |
+
"learning_rate": 4.578384798099763e-07,
|
| 510 |
+
"logits/chosen": -0.0849609375,
|
| 511 |
+
"logits/rejected": -0.515625,
|
| 512 |
+
"logps/chosen": -338.0,
|
| 513 |
+
"logps/rejected": -328.0,
|
| 514 |
+
"loss": 0.2903,
|
| 515 |
+
"rewards/accuracies": 0.925000011920929,
|
| 516 |
+
"rewards/chosen": 1.6953125,
|
| 517 |
+
"rewards/margins": 2.859375,
|
| 518 |
+
"rewards/rejected": -1.171875,
|
| 519 |
+
"step": 330
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 0.5448717948717948,
|
| 523 |
+
"grad_norm": 65.37858577012295,
|
| 524 |
+
"learning_rate": 4.548693586698337e-07,
|
| 525 |
+
"logits/chosen": -0.11083984375,
|
| 526 |
+
"logits/rejected": -0.28125,
|
| 527 |
+
"logps/chosen": -382.0,
|
| 528 |
+
"logps/rejected": -338.0,
|
| 529 |
+
"loss": 0.3046,
|
| 530 |
+
"rewards/accuracies": 0.862500011920929,
|
| 531 |
+
"rewards/chosen": 1.7421875,
|
| 532 |
+
"rewards/margins": 3.15625,
|
| 533 |
+
"rewards/rejected": -1.4140625,
|
| 534 |
+
"step": 340
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.5608974358974359,
|
| 538 |
+
"grad_norm": 53.122076726933024,
|
| 539 |
+
"learning_rate": 4.519002375296912e-07,
|
| 540 |
+
"logits/chosen": -0.1123046875,
|
| 541 |
+
"logits/rejected": -0.46875,
|
| 542 |
+
"logps/chosen": -370.0,
|
| 543 |
+
"logps/rejected": -358.0,
|
| 544 |
+
"loss": 0.3242,
|
| 545 |
+
"rewards/accuracies": 0.8500000238418579,
|
| 546 |
+
"rewards/chosen": 1.390625,
|
| 547 |
+
"rewards/margins": 2.765625,
|
| 548 |
+
"rewards/rejected": -1.375,
|
| 549 |
+
"step": 350
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 0.5769230769230769,
|
| 553 |
+
"grad_norm": 29.215237884321418,
|
| 554 |
+
"learning_rate": 4.4893111638954866e-07,
|
| 555 |
+
"logits/chosen": -0.267578125,
|
| 556 |
+
"logits/rejected": -0.3515625,
|
| 557 |
+
"logps/chosen": -380.0,
|
| 558 |
+
"logps/rejected": -324.0,
|
| 559 |
+
"loss": 0.1967,
|
| 560 |
+
"rewards/accuracies": 0.9375,
|
| 561 |
+
"rewards/chosen": 1.5390625,
|
| 562 |
+
"rewards/margins": 3.296875,
|
| 563 |
+
"rewards/rejected": -1.7578125,
|
| 564 |
+
"step": 360
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 0.592948717948718,
|
| 568 |
+
"grad_norm": 23.99951739667445,
|
| 569 |
+
"learning_rate": 4.4596199524940614e-07,
|
| 570 |
+
"logits/chosen": -0.24609375,
|
| 571 |
+
"logits/rejected": -0.376953125,
|
| 572 |
+
"logps/chosen": -392.0,
|
| 573 |
+
"logps/rejected": -328.0,
|
| 574 |
+
"loss": 0.2886,
|
| 575 |
+
"rewards/accuracies": 0.824999988079071,
|
| 576 |
+
"rewards/chosen": 1.0234375,
|
| 577 |
+
"rewards/margins": 2.515625,
|
| 578 |
+
"rewards/rejected": -1.5,
|
| 579 |
+
"step": 370
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"epoch": 0.6089743589743589,
|
| 583 |
+
"grad_norm": 73.48089183856715,
|
| 584 |
+
"learning_rate": 4.429928741092636e-07,
|
| 585 |
+
"logits/chosen": -0.1640625,
|
| 586 |
+
"logits/rejected": -0.443359375,
|
| 587 |
+
"logps/chosen": -378.0,
|
| 588 |
+
"logps/rejected": -346.0,
|
| 589 |
+
"loss": 0.2243,
|
| 590 |
+
"rewards/accuracies": 0.9125000238418579,
|
| 591 |
+
"rewards/chosen": 1.53125,
|
| 592 |
+
"rewards/margins": 3.125,
|
| 593 |
+
"rewards/rejected": -1.59375,
|
| 594 |
+
"step": 380
|
| 595 |
+
},
|
| 596 |
+
{
|
| 597 |
+
"epoch": 0.625,
|
| 598 |
+
"grad_norm": 34.2482977900771,
|
| 599 |
+
"learning_rate": 4.4002375296912115e-07,
|
| 600 |
+
"logits/chosen": -0.109375,
|
| 601 |
+
"logits/rejected": -0.5625,
|
| 602 |
+
"logps/chosen": -382.0,
|
| 603 |
+
"logps/rejected": -344.0,
|
| 604 |
+
"loss": 0.1972,
|
| 605 |
+
"rewards/accuracies": 0.8999999761581421,
|
| 606 |
+
"rewards/chosen": 1.46875,
|
| 607 |
+
"rewards/margins": 3.09375,
|
| 608 |
+
"rewards/rejected": -1.6328125,
|
| 609 |
+
"step": 390
|
| 610 |
+
},
|
| 611 |
+
{
|
| 612 |
+
"epoch": 0.6410256410256411,
|
| 613 |
+
"grad_norm": 46.4216688916151,
|
| 614 |
+
"learning_rate": 4.3705463182897863e-07,
|
| 615 |
+
"logits/chosen": -0.1435546875,
|
| 616 |
+
"logits/rejected": -0.3359375,
|
| 617 |
+
"logps/chosen": -372.0,
|
| 618 |
+
"logps/rejected": -340.0,
|
| 619 |
+
"loss": 0.2644,
|
| 620 |
+
"rewards/accuracies": 0.949999988079071,
|
| 621 |
+
"rewards/chosen": 2.265625,
|
| 622 |
+
"rewards/margins": 3.796875,
|
| 623 |
+
"rewards/rejected": -1.53125,
|
| 624 |
+
"step": 400
|
| 625 |
+
},
|
| 626 |
+
{
|
| 627 |
+
"epoch": 0.657051282051282,
|
| 628 |
+
"grad_norm": 44.3943999019337,
|
| 629 |
+
"learning_rate": 4.340855106888361e-07,
|
| 630 |
+
"logits/chosen": -0.1494140625,
|
| 631 |
+
"logits/rejected": -0.53515625,
|
| 632 |
+
"logps/chosen": -368.0,
|
| 633 |
+
"logps/rejected": -334.0,
|
| 634 |
+
"loss": 0.3043,
|
| 635 |
+
"rewards/accuracies": 0.9375,
|
| 636 |
+
"rewards/chosen": 1.6640625,
|
| 637 |
+
"rewards/margins": 3.265625,
|
| 638 |
+
"rewards/rejected": -1.59375,
|
| 639 |
+
"step": 410
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.6730769230769231,
|
| 643 |
+
"grad_norm": 49.38538485619709,
|
| 644 |
+
"learning_rate": 4.311163895486936e-07,
|
| 645 |
+
"logits/chosen": -0.2451171875,
|
| 646 |
+
"logits/rejected": -0.353515625,
|
| 647 |
+
"logps/chosen": -374.0,
|
| 648 |
+
"logps/rejected": -332.0,
|
| 649 |
+
"loss": 0.2839,
|
| 650 |
+
"rewards/accuracies": 0.7875000238418579,
|
| 651 |
+
"rewards/chosen": 1.5546875,
|
| 652 |
+
"rewards/margins": 2.96875,
|
| 653 |
+
"rewards/rejected": -1.4140625,
|
| 654 |
+
"step": 420
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 0.6891025641025641,
|
| 658 |
+
"grad_norm": 52.14918805667079,
|
| 659 |
+
"learning_rate": 4.28147268408551e-07,
|
| 660 |
+
"logits/chosen": -0.142578125,
|
| 661 |
+
"logits/rejected": -0.265625,
|
| 662 |
+
"logps/chosen": -392.0,
|
| 663 |
+
"logps/rejected": -308.0,
|
| 664 |
+
"loss": 0.2573,
|
| 665 |
+
"rewards/accuracies": 0.8500000238418579,
|
| 666 |
+
"rewards/chosen": 1.71875,
|
| 667 |
+
"rewards/margins": 3.546875,
|
| 668 |
+
"rewards/rejected": -1.8359375,
|
| 669 |
+
"step": 430
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"epoch": 0.7051282051282052,
|
| 673 |
+
"grad_norm": 41.67427154859993,
|
| 674 |
+
"learning_rate": 4.251781472684085e-07,
|
| 675 |
+
"logits/chosen": -0.2041015625,
|
| 676 |
+
"logits/rejected": -0.19140625,
|
| 677 |
+
"logps/chosen": -382.0,
|
| 678 |
+
"logps/rejected": -302.0,
|
| 679 |
+
"loss": 0.2207,
|
| 680 |
+
"rewards/accuracies": 0.8999999761581421,
|
| 681 |
+
"rewards/chosen": 1.9609375,
|
| 682 |
+
"rewards/margins": 2.921875,
|
| 683 |
+
"rewards/rejected": -0.96484375,
|
| 684 |
+
"step": 440
|
| 685 |
+
},
|
| 686 |
+
{
|
| 687 |
+
"epoch": 0.7211538461538461,
|
| 688 |
+
"grad_norm": 49.967389161188926,
|
| 689 |
+
"learning_rate": 4.22209026128266e-07,
|
| 690 |
+
"logits/chosen": -0.216796875,
|
| 691 |
+
"logits/rejected": -0.33984375,
|
| 692 |
+
"logps/chosen": -368.0,
|
| 693 |
+
"logps/rejected": -316.0,
|
| 694 |
+
"loss": 0.3459,
|
| 695 |
+
"rewards/accuracies": 0.875,
|
| 696 |
+
"rewards/chosen": 1.46875,
|
| 697 |
+
"rewards/margins": 2.75,
|
| 698 |
+
"rewards/rejected": -1.28125,
|
| 699 |
+
"step": 450
|
| 700 |
+
},
|
| 701 |
+
{
|
| 702 |
+
"epoch": 0.7371794871794872,
|
| 703 |
+
"grad_norm": 45.29394404835159,
|
| 704 |
+
"learning_rate": 4.192399049881235e-07,
|
| 705 |
+
"logits/chosen": -0.259765625,
|
| 706 |
+
"logits/rejected": -0.462890625,
|
| 707 |
+
"logps/chosen": -392.0,
|
| 708 |
+
"logps/rejected": -334.0,
|
| 709 |
+
"loss": 0.2416,
|
| 710 |
+
"rewards/accuracies": 0.9125000238418579,
|
| 711 |
+
"rewards/chosen": 1.765625,
|
| 712 |
+
"rewards/margins": 3.296875,
|
| 713 |
+
"rewards/rejected": -1.53125,
|
| 714 |
+
"step": 460
|
| 715 |
+
},
|
| 716 |
+
{
|
| 717 |
+
"epoch": 0.7532051282051282,
|
| 718 |
+
"grad_norm": 49.65359950033634,
|
| 719 |
+
"learning_rate": 4.16270783847981e-07,
|
| 720 |
+
"logits/chosen": -0.09814453125,
|
| 721 |
+
"logits/rejected": -0.384765625,
|
| 722 |
+
"logps/chosen": -340.0,
|
| 723 |
+
"logps/rejected": -312.0,
|
| 724 |
+
"loss": 0.2871,
|
| 725 |
+
"rewards/accuracies": 0.8999999761581421,
|
| 726 |
+
"rewards/chosen": 2.25,
|
| 727 |
+
"rewards/margins": 3.34375,
|
| 728 |
+
"rewards/rejected": -1.09375,
|
| 729 |
+
"step": 470
|
| 730 |
+
},
|
| 731 |
+
{
|
| 732 |
+
"epoch": 0.7692307692307693,
|
| 733 |
+
"grad_norm": 34.79083088842419,
|
| 734 |
+
"learning_rate": 4.1330166270783846e-07,
|
| 735 |
+
"logits/chosen": -0.1796875,
|
| 736 |
+
"logits/rejected": -0.423828125,
|
| 737 |
+
"logps/chosen": -358.0,
|
| 738 |
+
"logps/rejected": -344.0,
|
| 739 |
+
"loss": 0.2121,
|
| 740 |
+
"rewards/accuracies": 0.925000011920929,
|
| 741 |
+
"rewards/chosen": 1.453125,
|
| 742 |
+
"rewards/margins": 3.25,
|
| 743 |
+
"rewards/rejected": -1.8046875,
|
| 744 |
+
"step": 480
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 0.7852564102564102,
|
| 748 |
+
"grad_norm": 24.646692522257364,
|
| 749 |
+
"learning_rate": 4.1033254156769594e-07,
|
| 750 |
+
"logits/chosen": -0.1435546875,
|
| 751 |
+
"logits/rejected": -0.390625,
|
| 752 |
+
"logps/chosen": -360.0,
|
| 753 |
+
"logps/rejected": -332.0,
|
| 754 |
+
"loss": 0.2359,
|
| 755 |
+
"rewards/accuracies": 0.9125000238418579,
|
| 756 |
+
"rewards/chosen": 1.6640625,
|
| 757 |
+
"rewards/margins": 3.421875,
|
| 758 |
+
"rewards/rejected": -1.75,
|
| 759 |
+
"step": 490
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 0.8012820512820513,
|
| 763 |
+
"grad_norm": 32.88069299074813,
|
| 764 |
+
"learning_rate": 4.0736342042755347e-07,
|
| 765 |
+
"logits/chosen": -0.216796875,
|
| 766 |
+
"logits/rejected": -0.5390625,
|
| 767 |
+
"logps/chosen": -380.0,
|
| 768 |
+
"logps/rejected": -328.0,
|
| 769 |
+
"loss": 0.2243,
|
| 770 |
+
"rewards/accuracies": 0.8500000238418579,
|
| 771 |
+
"rewards/chosen": 0.9609375,
|
| 772 |
+
"rewards/margins": 2.90625,
|
| 773 |
+
"rewards/rejected": -1.9453125,
|
| 774 |
+
"step": 500
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 0.8173076923076923,
|
| 778 |
+
"grad_norm": 35.40990218888171,
|
| 779 |
+
"learning_rate": 4.0439429928741095e-07,
|
| 780 |
+
"logits/chosen": -0.173828125,
|
| 781 |
+
"logits/rejected": -0.302734375,
|
| 782 |
+
"logps/chosen": -388.0,
|
| 783 |
+
"logps/rejected": -320.0,
|
| 784 |
+
"loss": 0.2403,
|
| 785 |
+
"rewards/accuracies": 0.887499988079071,
|
| 786 |
+
"rewards/chosen": 1.421875,
|
| 787 |
+
"rewards/margins": 3.203125,
|
| 788 |
+
"rewards/rejected": -1.78125,
|
| 789 |
+
"step": 510
|
| 790 |
+
},
|
| 791 |
+
{
|
| 792 |
+
"epoch": 0.8333333333333334,
|
| 793 |
+
"grad_norm": 45.40116394651564,
|
| 794 |
+
"learning_rate": 4.0142517814726837e-07,
|
| 795 |
+
"logits/chosen": -0.220703125,
|
| 796 |
+
"logits/rejected": -0.37109375,
|
| 797 |
+
"logps/chosen": -398.0,
|
| 798 |
+
"logps/rejected": -340.0,
|
| 799 |
+
"loss": 0.2562,
|
| 800 |
+
"rewards/accuracies": 0.875,
|
| 801 |
+
"rewards/chosen": 1.4375,
|
| 802 |
+
"rewards/margins": 3.0,
|
| 803 |
+
"rewards/rejected": -1.5625,
|
| 804 |
+
"step": 520
|
| 805 |
+
},
|
| 806 |
+
{
|
| 807 |
+
"epoch": 0.8493589743589743,
|
| 808 |
+
"grad_norm": 38.39116264620735,
|
| 809 |
+
"learning_rate": 3.9845605700712585e-07,
|
| 810 |
+
"logits/chosen": -0.1953125,
|
| 811 |
+
"logits/rejected": -0.244140625,
|
| 812 |
+
"logps/chosen": -374.0,
|
| 813 |
+
"logps/rejected": -288.0,
|
| 814 |
+
"loss": 0.2727,
|
| 815 |
+
"rewards/accuracies": 0.875,
|
| 816 |
+
"rewards/chosen": 1.3671875,
|
| 817 |
+
"rewards/margins": 3.046875,
|
| 818 |
+
"rewards/rejected": -1.6875,
|
| 819 |
+
"step": 530
|
| 820 |
+
},
|
| 821 |
+
{
|
| 822 |
+
"epoch": 0.8653846153846154,
|
| 823 |
+
"grad_norm": 47.38780960867294,
|
| 824 |
+
"learning_rate": 3.9548693586698333e-07,
|
| 825 |
+
"logits/chosen": -0.24609375,
|
| 826 |
+
"logits/rejected": -0.4140625,
|
| 827 |
+
"logps/chosen": -372.0,
|
| 828 |
+
"logps/rejected": -318.0,
|
| 829 |
+
"loss": 0.2017,
|
| 830 |
+
"rewards/accuracies": 0.949999988079071,
|
| 831 |
+
"rewards/chosen": 1.9296875,
|
| 832 |
+
"rewards/margins": 3.421875,
|
| 833 |
+
"rewards/rejected": -1.4921875,
|
| 834 |
+
"step": 540
|
| 835 |
+
},
|
| 836 |
+
{
|
| 837 |
+
"epoch": 0.8814102564102564,
|
| 838 |
+
"grad_norm": 53.98621268772635,
|
| 839 |
+
"learning_rate": 3.925178147268408e-07,
|
| 840 |
+
"logits/chosen": -0.322265625,
|
| 841 |
+
"logits/rejected": -0.283203125,
|
| 842 |
+
"logps/chosen": -358.0,
|
| 843 |
+
"logps/rejected": -314.0,
|
| 844 |
+
"loss": 0.2358,
|
| 845 |
+
"rewards/accuracies": 0.8999999761581421,
|
| 846 |
+
"rewards/chosen": 2.078125,
|
| 847 |
+
"rewards/margins": 3.234375,
|
| 848 |
+
"rewards/rejected": -1.1640625,
|
| 849 |
+
"step": 550
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 0.8974358974358975,
|
| 853 |
+
"grad_norm": 57.737496610523515,
|
| 854 |
+
"learning_rate": 3.8954869358669834e-07,
|
| 855 |
+
"logits/chosen": -0.1767578125,
|
| 856 |
+
"logits/rejected": -0.271484375,
|
| 857 |
+
"logps/chosen": -384.0,
|
| 858 |
+
"logps/rejected": -332.0,
|
| 859 |
+
"loss": 0.2313,
|
| 860 |
+
"rewards/accuracies": 0.875,
|
| 861 |
+
"rewards/chosen": 1.8125,
|
| 862 |
+
"rewards/margins": 3.203125,
|
| 863 |
+
"rewards/rejected": -1.3828125,
|
| 864 |
+
"step": 560
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 0.9134615384615384,
|
| 868 |
+
"grad_norm": 29.929989102081784,
|
| 869 |
+
"learning_rate": 3.865795724465558e-07,
|
| 870 |
+
"logits/chosen": -0.26171875,
|
| 871 |
+
"logits/rejected": -0.466796875,
|
| 872 |
+
"logps/chosen": -366.0,
|
| 873 |
+
"logps/rejected": -302.0,
|
| 874 |
+
"loss": 0.2186,
|
| 875 |
+
"rewards/accuracies": 0.887499988079071,
|
| 876 |
+
"rewards/chosen": 2.0,
|
| 877 |
+
"rewards/margins": 3.78125,
|
| 878 |
+
"rewards/rejected": -1.78125,
|
| 879 |
+
"step": 570
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"epoch": 0.9294871794871795,
|
| 883 |
+
"grad_norm": 14.744590527736637,
|
| 884 |
+
"learning_rate": 3.836104513064133e-07,
|
| 885 |
+
"logits/chosen": -0.30078125,
|
| 886 |
+
"logits/rejected": -0.46875,
|
| 887 |
+
"logps/chosen": -364.0,
|
| 888 |
+
"logps/rejected": -324.0,
|
| 889 |
+
"loss": 0.1895,
|
| 890 |
+
"rewards/accuracies": 0.949999988079071,
|
| 891 |
+
"rewards/chosen": 1.84375,
|
| 892 |
+
"rewards/margins": 3.75,
|
| 893 |
+
"rewards/rejected": -1.9140625,
|
| 894 |
+
"step": 580
|
| 895 |
+
},
|
| 896 |
+
{
|
| 897 |
+
"epoch": 0.9455128205128205,
|
| 898 |
+
"grad_norm": 43.4367923218673,
|
| 899 |
+
"learning_rate": 3.806413301662708e-07,
|
| 900 |
+
"logits/chosen": -0.154296875,
|
| 901 |
+
"logits/rejected": -0.5,
|
| 902 |
+
"logps/chosen": -366.0,
|
| 903 |
+
"logps/rejected": -348.0,
|
| 904 |
+
"loss": 0.2121,
|
| 905 |
+
"rewards/accuracies": 0.875,
|
| 906 |
+
"rewards/chosen": 1.625,
|
| 907 |
+
"rewards/margins": 3.546875,
|
| 908 |
+
"rewards/rejected": -1.9140625,
|
| 909 |
+
"step": 590
|
| 910 |
+
},
|
| 911 |
+
{
|
| 912 |
+
"epoch": 0.9615384615384616,
|
| 913 |
+
"grad_norm": 14.598190302159233,
|
| 914 |
+
"learning_rate": 3.7767220902612825e-07,
|
| 915 |
+
"logits/chosen": -0.21875,
|
| 916 |
+
"logits/rejected": -0.35546875,
|
| 917 |
+
"logps/chosen": -374.0,
|
| 918 |
+
"logps/rejected": -328.0,
|
| 919 |
+
"loss": 0.2134,
|
| 920 |
+
"rewards/accuracies": 0.8999999761581421,
|
| 921 |
+
"rewards/chosen": 1.4609375,
|
| 922 |
+
"rewards/margins": 3.59375,
|
| 923 |
+
"rewards/rejected": -2.140625,
|
| 924 |
+
"step": 600
|
| 925 |
+
},
|
| 926 |
+
{
|
| 927 |
+
"epoch": 0.9775641025641025,
|
| 928 |
+
"grad_norm": 65.50233470122603,
|
| 929 |
+
"learning_rate": 3.747030878859858e-07,
|
| 930 |
+
"logits/chosen": -0.09375,
|
| 931 |
+
"logits/rejected": -0.36328125,
|
| 932 |
+
"logps/chosen": -348.0,
|
| 933 |
+
"logps/rejected": -332.0,
|
| 934 |
+
"loss": 0.3465,
|
| 935 |
+
"rewards/accuracies": 0.887499988079071,
|
| 936 |
+
"rewards/chosen": 1.484375,
|
| 937 |
+
"rewards/margins": 3.65625,
|
| 938 |
+
"rewards/rejected": -2.171875,
|
| 939 |
+
"step": 610
|
| 940 |
+
},
|
| 941 |
+
{
|
| 942 |
+
"epoch": 0.9935897435897436,
|
| 943 |
+
"grad_norm": 45.476213003555294,
|
| 944 |
+
"learning_rate": 3.717339667458432e-07,
|
| 945 |
+
"logits/chosen": -0.158203125,
|
| 946 |
+
"logits/rejected": -0.439453125,
|
| 947 |
+
"logps/chosen": -364.0,
|
| 948 |
+
"logps/rejected": -330.0,
|
| 949 |
+
"loss": 0.2557,
|
| 950 |
+
"rewards/accuracies": 0.875,
|
| 951 |
+
"rewards/chosen": 1.1796875,
|
| 952 |
+
"rewards/margins": 3.140625,
|
| 953 |
+
"rewards/rejected": -1.953125,
|
| 954 |
+
"step": 620
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 1.0,
|
| 958 |
+
"eval_logits/chosen": -0.2333984375,
|
| 959 |
+
"eval_logits/rejected": -0.328125,
|
| 960 |
+
"eval_logps/chosen": -366.0,
|
| 961 |
+
"eval_logps/rejected": -340.0,
|
| 962 |
+
"eval_loss": 0.20789062976837158,
|
| 963 |
+
"eval_rewards/accuracies": 0.8942307829856873,
|
| 964 |
+
"eval_rewards/chosen": 1.5859375,
|
| 965 |
+
"eval_rewards/margins": 3.375,
|
| 966 |
+
"eval_rewards/rejected": -1.7890625,
|
| 967 |
+
"eval_runtime": 28.0659,
|
| 968 |
+
"eval_samples_per_second": 7.126,
|
| 969 |
+
"eval_steps_per_second": 0.463,
|
| 970 |
+
"step": 624
|
| 971 |
+
},
|
| 972 |
+
{
|
| 973 |
+
"epoch": 1.0096153846153846,
|
| 974 |
+
"grad_norm": 6.092991588272169,
|
| 975 |
+
"learning_rate": 3.687648456057007e-07,
|
| 976 |
+
"logits/chosen": -0.2734375,
|
| 977 |
+
"logits/rejected": -0.388671875,
|
| 978 |
+
"logps/chosen": -350.0,
|
| 979 |
+
"logps/rejected": -322.0,
|
| 980 |
+
"loss": 0.1514,
|
| 981 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 982 |
+
"rewards/chosen": 1.7734375,
|
| 983 |
+
"rewards/margins": 3.96875,
|
| 984 |
+
"rewards/rejected": -2.203125,
|
| 985 |
+
"step": 630
|
| 986 |
+
},
|
| 987 |
+
{
|
| 988 |
+
"epoch": 1.0256410256410255,
|
| 989 |
+
"grad_norm": 15.11610385237531,
|
| 990 |
+
"learning_rate": 3.6579572446555817e-07,
|
| 991 |
+
"logits/chosen": -0.21875,
|
| 992 |
+
"logits/rejected": -0.46484375,
|
| 993 |
+
"logps/chosen": -356.0,
|
| 994 |
+
"logps/rejected": -374.0,
|
| 995 |
+
"loss": 0.0783,
|
| 996 |
+
"rewards/accuracies": 0.987500011920929,
|
| 997 |
+
"rewards/chosen": 1.921875,
|
| 998 |
+
"rewards/margins": 4.8125,
|
| 999 |
+
"rewards/rejected": -2.890625,
|
| 1000 |
+
"step": 640
|
| 1001 |
+
},
|
| 1002 |
+
{
|
| 1003 |
+
"epoch": 1.0416666666666667,
|
| 1004 |
+
"grad_norm": 8.924081971380067,
|
| 1005 |
+
"learning_rate": 3.6282660332541565e-07,
|
| 1006 |
+
"logits/chosen": -0.36328125,
|
| 1007 |
+
"logits/rejected": -0.5,
|
| 1008 |
+
"logps/chosen": -378.0,
|
| 1009 |
+
"logps/rejected": -348.0,
|
| 1010 |
+
"loss": 0.0486,
|
| 1011 |
+
"rewards/accuracies": 1.0,
|
| 1012 |
+
"rewards/chosen": 1.7421875,
|
| 1013 |
+
"rewards/margins": 4.59375,
|
| 1014 |
+
"rewards/rejected": -2.84375,
|
| 1015 |
+
"step": 650
|
| 1016 |
+
},
|
| 1017 |
+
{
|
| 1018 |
+
"epoch": 1.0576923076923077,
|
| 1019 |
+
"grad_norm": 8.969543433658254,
|
| 1020 |
+
"learning_rate": 3.598574821852731e-07,
|
| 1021 |
+
"logits/chosen": -0.1572265625,
|
| 1022 |
+
"logits/rejected": -0.2890625,
|
| 1023 |
+
"logps/chosen": -364.0,
|
| 1024 |
+
"logps/rejected": -344.0,
|
| 1025 |
+
"loss": 0.1064,
|
| 1026 |
+
"rewards/accuracies": 0.949999988079071,
|
| 1027 |
+
"rewards/chosen": 2.109375,
|
| 1028 |
+
"rewards/margins": 4.1875,
|
| 1029 |
+
"rewards/rejected": -2.078125,
|
| 1030 |
+
"step": 660
|
| 1031 |
+
},
|
| 1032 |
+
{
|
| 1033 |
+
"epoch": 1.0737179487179487,
|
| 1034 |
+
"grad_norm": 20.194350083453877,
|
| 1035 |
+
"learning_rate": 3.5688836104513066e-07,
|
| 1036 |
+
"logits/chosen": -0.1884765625,
|
| 1037 |
+
"logits/rejected": -0.427734375,
|
| 1038 |
+
"logps/chosen": -376.0,
|
| 1039 |
+
"logps/rejected": -326.0,
|
| 1040 |
+
"loss": 0.0802,
|
| 1041 |
+
"rewards/accuracies": 0.987500011920929,
|
| 1042 |
+
"rewards/chosen": 2.109375,
|
| 1043 |
+
"rewards/margins": 4.3125,
|
| 1044 |
+
"rewards/rejected": -2.1875,
|
| 1045 |
+
"step": 670
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 1.0897435897435896,
|
| 1049 |
+
"grad_norm": 9.472914566833554,
|
| 1050 |
+
"learning_rate": 3.5391923990498813e-07,
|
| 1051 |
+
"logits/chosen": -0.1494140625,
|
| 1052 |
+
"logits/rejected": -0.39453125,
|
| 1053 |
+
"logps/chosen": -354.0,
|
| 1054 |
+
"logps/rejected": -330.0,
|
| 1055 |
+
"loss": 0.0748,
|
| 1056 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 1057 |
+
"rewards/chosen": 2.671875,
|
| 1058 |
+
"rewards/margins": 5.0625,
|
| 1059 |
+
"rewards/rejected": -2.390625,
|
| 1060 |
+
"step": 680
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 1.1057692307692308,
|
| 1064 |
+
"grad_norm": 27.091485921309573,
|
| 1065 |
+
"learning_rate": 3.509501187648456e-07,
|
| 1066 |
+
"logits/chosen": -0.2578125,
|
| 1067 |
+
"logits/rejected": -0.359375,
|
| 1068 |
+
"logps/chosen": -368.0,
|
| 1069 |
+
"logps/rejected": -324.0,
|
| 1070 |
+
"loss": 0.0862,
|
| 1071 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 1072 |
+
"rewards/chosen": 2.140625,
|
| 1073 |
+
"rewards/margins": 4.71875,
|
| 1074 |
+
"rewards/rejected": -2.5625,
|
| 1075 |
+
"step": 690
|
| 1076 |
+
},
|
| 1077 |
+
{
|
| 1078 |
+
"epoch": 1.1217948717948718,
|
| 1079 |
+
"grad_norm": 23.55129242956029,
|
| 1080 |
+
"learning_rate": 3.479809976247031e-07,
|
| 1081 |
+
"logits/chosen": -0.1875,
|
| 1082 |
+
"logits/rejected": -0.25,
|
| 1083 |
+
"logps/chosen": -352.0,
|
| 1084 |
+
"logps/rejected": -314.0,
|
| 1085 |
+
"loss": 0.0667,
|
| 1086 |
+
"rewards/accuracies": 0.987500011920929,
|
| 1087 |
+
"rewards/chosen": 2.21875,
|
| 1088 |
+
"rewards/margins": 4.65625,
|
| 1089 |
+
"rewards/rejected": -2.4375,
|
| 1090 |
+
"step": 700
|
| 1091 |
+
},
|
| 1092 |
+
{
|
| 1093 |
+
"epoch": 1.1378205128205128,
|
| 1094 |
+
"grad_norm": 16.191916223045457,
|
| 1095 |
+
"learning_rate": 3.450118764845605e-07,
|
| 1096 |
+
"logits/chosen": -0.2431640625,
|
| 1097 |
+
"logits/rejected": -0.4453125,
|
| 1098 |
+
"logps/chosen": -368.0,
|
| 1099 |
+
"logps/rejected": -352.0,
|
| 1100 |
+
"loss": 0.1058,
|
| 1101 |
+
"rewards/accuracies": 0.9375,
|
| 1102 |
+
"rewards/chosen": 1.953125,
|
| 1103 |
+
"rewards/margins": 4.84375,
|
| 1104 |
+
"rewards/rejected": -2.890625,
|
| 1105 |
+
"step": 710
|
| 1106 |
+
},
|
| 1107 |
+
{
|
| 1108 |
+
"epoch": 1.1538461538461537,
|
| 1109 |
+
"grad_norm": 6.197082349513418,
|
| 1110 |
+
"learning_rate": 3.42042755344418e-07,
|
| 1111 |
+
"logits/chosen": -0.193359375,
|
| 1112 |
+
"logits/rejected": -0.298828125,
|
| 1113 |
+
"logps/chosen": -420.0,
|
| 1114 |
+
"logps/rejected": -352.0,
|
| 1115 |
+
"loss": 0.1152,
|
| 1116 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 1117 |
+
"rewards/chosen": 1.140625,
|
| 1118 |
+
"rewards/margins": 4.8125,
|
| 1119 |
+
"rewards/rejected": -3.65625,
|
| 1120 |
+
"step": 720
|
| 1121 |
+
},
|
| 1122 |
+
{
|
| 1123 |
+
"epoch": 1.169871794871795,
|
| 1124 |
+
"grad_norm": 20.175263771423996,
|
| 1125 |
+
"learning_rate": 3.390736342042755e-07,
|
| 1126 |
+
"logits/chosen": -0.0849609375,
|
| 1127 |
+
"logits/rejected": -0.30859375,
|
| 1128 |
+
"logps/chosen": -368.0,
|
| 1129 |
+
"logps/rejected": -340.0,
|
| 1130 |
+
"loss": 0.0717,
|
| 1131 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 1132 |
+
"rewards/chosen": 1.8203125,
|
| 1133 |
+
"rewards/margins": 5.125,
|
| 1134 |
+
"rewards/rejected": -3.296875,
|
| 1135 |
+
"step": 730
|
| 1136 |
+
},
|
| 1137 |
+
{
|
| 1138 |
+
"epoch": 1.185897435897436,
|
| 1139 |
+
"grad_norm": 14.053624440757009,
|
| 1140 |
+
"learning_rate": 3.36104513064133e-07,
|
| 1141 |
+
"logits/chosen": -0.1591796875,
|
| 1142 |
+
"logits/rejected": -0.373046875,
|
| 1143 |
+
"logps/chosen": -368.0,
|
| 1144 |
+
"logps/rejected": -342.0,
|
| 1145 |
+
"loss": 0.0763,
|
| 1146 |
+
"rewards/accuracies": 0.987500011920929,
|
| 1147 |
+
"rewards/chosen": 2.0,
|
| 1148 |
+
"rewards/margins": 5.53125,
|
| 1149 |
+
"rewards/rejected": -3.515625,
|
| 1150 |
+
"step": 740
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 1.2019230769230769,
|
| 1154 |
+
"grad_norm": 19.07797922710506,
|
| 1155 |
+
"learning_rate": 3.331353919239905e-07,
|
| 1156 |
+
"logits/chosen": -0.134765625,
|
| 1157 |
+
"logits/rejected": -0.37109375,
|
| 1158 |
+
"logps/chosen": -374.0,
|
| 1159 |
+
"logps/rejected": -346.0,
|
| 1160 |
+
"loss": 0.0837,
|
| 1161 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 1162 |
+
"rewards/chosen": 1.859375,
|
| 1163 |
+
"rewards/margins": 4.9375,
|
| 1164 |
+
"rewards/rejected": -3.078125,
|
| 1165 |
+
"step": 750
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 1.217948717948718,
|
| 1169 |
+
"grad_norm": 3.621341208921815,
|
| 1170 |
+
"learning_rate": 3.3016627078384796e-07,
|
| 1171 |
+
"logits/chosen": -0.1337890625,
|
| 1172 |
+
"logits/rejected": -0.447265625,
|
| 1173 |
+
"logps/chosen": -382.0,
|
| 1174 |
+
"logps/rejected": -362.0,
|
| 1175 |
+
"loss": 0.0574,
|
| 1176 |
+
"rewards/accuracies": 0.949999988079071,
|
| 1177 |
+
"rewards/chosen": 1.7734375,
|
| 1178 |
+
"rewards/margins": 5.0625,
|
| 1179 |
+
"rewards/rejected": -3.28125,
|
| 1180 |
+
"step": 760
|
| 1181 |
+
},
|
| 1182 |
+
{
|
| 1183 |
+
"epoch": 1.233974358974359,
|
| 1184 |
+
"grad_norm": 32.55125289195706,
|
| 1185 |
+
"learning_rate": 3.2719714964370544e-07,
|
| 1186 |
+
"logits/chosen": -0.2099609375,
|
| 1187 |
+
"logits/rejected": -0.408203125,
|
| 1188 |
+
"logps/chosen": -354.0,
|
| 1189 |
+
"logps/rejected": -330.0,
|
| 1190 |
+
"loss": 0.0801,
|
| 1191 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 1192 |
+
"rewards/chosen": 2.203125,
|
| 1193 |
+
"rewards/margins": 5.6875,
|
| 1194 |
+
"rewards/rejected": -3.46875,
|
| 1195 |
+
"step": 770
|
| 1196 |
+
},
|
| 1197 |
+
{
|
| 1198 |
+
"epoch": 1.25,
|
| 1199 |
+
"grad_norm": 22.82618457101383,
|
| 1200 |
+
"learning_rate": 3.2422802850356297e-07,
|
| 1201 |
+
"logits/chosen": -0.265625,
|
| 1202 |
+
"logits/rejected": -0.486328125,
|
| 1203 |
+
"logps/chosen": -362.0,
|
| 1204 |
+
"logps/rejected": -334.0,
|
| 1205 |
+
"loss": 0.0937,
|
| 1206 |
+
"rewards/accuracies": 0.925000011920929,
|
| 1207 |
+
"rewards/chosen": 2.03125,
|
| 1208 |
+
"rewards/margins": 5.0625,
|
| 1209 |
+
"rewards/rejected": -3.046875,
|
| 1210 |
+
"step": 780
|
| 1211 |
+
},
|
| 1212 |
+
{
|
| 1213 |
+
"epoch": 1.266025641025641,
|
| 1214 |
+
"grad_norm": 49.835975518914886,
|
| 1215 |
+
"learning_rate": 3.2125890736342045e-07,
|
| 1216 |
+
"logits/chosen": -0.24609375,
|
| 1217 |
+
"logits/rejected": -0.447265625,
|
| 1218 |
+
"logps/chosen": -360.0,
|
| 1219 |
+
"logps/rejected": -352.0,
|
| 1220 |
+
"loss": 0.0787,
|
| 1221 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 1222 |
+
"rewards/chosen": 2.140625,
|
| 1223 |
+
"rewards/margins": 5.5,
|
| 1224 |
+
"rewards/rejected": -3.375,
|
| 1225 |
+
"step": 790
|
| 1226 |
+
},
|
| 1227 |
+
{
|
| 1228 |
+
"epoch": 1.282051282051282,
|
| 1229 |
+
"grad_norm": 3.1313003868772107,
|
| 1230 |
+
"learning_rate": 3.182897862232779e-07,
|
| 1231 |
+
"logits/chosen": -0.283203125,
|
| 1232 |
+
"logits/rejected": -0.287109375,
|
| 1233 |
+
"logps/chosen": -372.0,
|
| 1234 |
+
"logps/rejected": -326.0,
|
| 1235 |
+
"loss": 0.1086,
|
| 1236 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 1237 |
+
"rewards/chosen": 1.6171875,
|
| 1238 |
+
"rewards/margins": 4.78125,
|
| 1239 |
+
"rewards/rejected": -3.1875,
|
| 1240 |
+
"step": 800
|
| 1241 |
+
},
|
| 1242 |
+
{
|
| 1243 |
+
"epoch": 1.2980769230769231,
|
| 1244 |
+
"grad_norm": 13.950684325273025,
|
| 1245 |
+
"learning_rate": 3.1532066508313535e-07,
|
| 1246 |
+
"logits/chosen": -0.234375,
|
| 1247 |
+
"logits/rejected": -0.640625,
|
| 1248 |
+
"logps/chosen": -344.0,
|
| 1249 |
+
"logps/rejected": -344.0,
|
| 1250 |
+
"loss": 0.0843,
|
| 1251 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 1252 |
+
"rewards/chosen": 2.140625,
|
| 1253 |
+
"rewards/margins": 5.0625,
|
| 1254 |
+
"rewards/rejected": -2.90625,
|
| 1255 |
+
"step": 810
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 1.314102564102564,
|
| 1259 |
+
"grad_norm": 6.930473305635045,
|
| 1260 |
+
"learning_rate": 3.1235154394299283e-07,
|
| 1261 |
+
"logits/chosen": -0.341796875,
|
| 1262 |
+
"logits/rejected": -0.326171875,
|
| 1263 |
+
"logps/chosen": -362.0,
|
| 1264 |
+
"logps/rejected": -344.0,
|
| 1265 |
+
"loss": 0.0667,
|
| 1266 |
+
"rewards/accuracies": 0.987500011920929,
|
| 1267 |
+
"rewards/chosen": 1.984375,
|
| 1268 |
+
"rewards/margins": 5.21875,
|
| 1269 |
+
"rewards/rejected": -3.234375,
|
| 1270 |
+
"step": 820
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"epoch": 1.330128205128205,
|
| 1274 |
+
"grad_norm": 21.31500643130334,
|
| 1275 |
+
"learning_rate": 3.093824228028503e-07,
|
| 1276 |
+
"logits/chosen": -0.1748046875,
|
| 1277 |
+
"logits/rejected": -0.443359375,
|
| 1278 |
+
"logps/chosen": -366.0,
|
| 1279 |
+
"logps/rejected": -342.0,
|
| 1280 |
+
"loss": 0.0595,
|
| 1281 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 1282 |
+
"rewards/chosen": 1.984375,
|
| 1283 |
+
"rewards/margins": 5.5625,
|
| 1284 |
+
"rewards/rejected": -3.5625,
|
| 1285 |
+
"step": 830
|
| 1286 |
+
},
|
| 1287 |
+
{
|
| 1288 |
+
"epoch": 1.3461538461538463,
|
| 1289 |
+
"grad_norm": 32.89803571469975,
|
| 1290 |
+
"learning_rate": 3.0641330166270784e-07,
|
| 1291 |
+
"logits/chosen": -0.234375,
|
| 1292 |
+
"logits/rejected": -0.34375,
|
| 1293 |
+
"logps/chosen": -330.0,
|
| 1294 |
+
"logps/rejected": -374.0,
|
| 1295 |
+
"loss": 0.0511,
|
| 1296 |
+
"rewards/accuracies": 1.0,
|
| 1297 |
+
"rewards/chosen": 2.28125,
|
| 1298 |
+
"rewards/margins": 6.0625,
|
| 1299 |
+
"rewards/rejected": -3.8125,
|
| 1300 |
+
"step": 840
|
| 1301 |
+
},
|
| 1302 |
+
{
|
| 1303 |
+
"epoch": 1.3621794871794872,
|
| 1304 |
+
"grad_norm": 32.22922106524605,
|
| 1305 |
+
"learning_rate": 3.034441805225653e-07,
|
| 1306 |
+
"logits/chosen": -0.279296875,
|
| 1307 |
+
"logits/rejected": -0.32421875,
|
| 1308 |
+
"logps/chosen": -358.0,
|
| 1309 |
+
"logps/rejected": -358.0,
|
| 1310 |
+
"loss": 0.0406,
|
| 1311 |
+
"rewards/accuracies": 0.987500011920929,
|
| 1312 |
+
"rewards/chosen": 2.609375,
|
| 1313 |
+
"rewards/margins": 6.21875,
|
| 1314 |
+
"rewards/rejected": -3.59375,
|
| 1315 |
+
"step": 850
|
| 1316 |
+
},
|
| 1317 |
+
{
|
| 1318 |
+
"epoch": 1.3782051282051282,
|
| 1319 |
+
"grad_norm": 28.46125519366018,
|
| 1320 |
+
"learning_rate": 3.004750593824228e-07,
|
| 1321 |
+
"logits/chosen": -0.2294921875,
|
| 1322 |
+
"logits/rejected": -0.19140625,
|
| 1323 |
+
"logps/chosen": -348.0,
|
| 1324 |
+
"logps/rejected": -330.0,
|
| 1325 |
+
"loss": 0.0549,
|
| 1326 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 1327 |
+
"rewards/chosen": 2.109375,
|
| 1328 |
+
"rewards/margins": 6.15625,
|
| 1329 |
+
"rewards/rejected": -4.0625,
|
| 1330 |
+
"step": 860
|
| 1331 |
+
},
|
| 1332 |
+
{
|
| 1333 |
+
"epoch": 1.3942307692307692,
|
| 1334 |
+
"grad_norm": 14.047872703813512,
|
| 1335 |
+
"learning_rate": 2.975059382422803e-07,
|
| 1336 |
+
"logits/chosen": -0.2490234375,
|
| 1337 |
+
"logits/rejected": -0.359375,
|
| 1338 |
+
"logps/chosen": -366.0,
|
| 1339 |
+
"logps/rejected": -316.0,
|
| 1340 |
+
"loss": 0.0797,
|
| 1341 |
+
"rewards/accuracies": 0.949999988079071,
|
| 1342 |
+
"rewards/chosen": 2.296875,
|
| 1343 |
+
"rewards/margins": 6.375,
|
| 1344 |
+
"rewards/rejected": -4.09375,
|
| 1345 |
+
"step": 870
|
| 1346 |
+
},
|
| 1347 |
+
{
|
| 1348 |
+
"epoch": 1.4102564102564101,
|
| 1349 |
+
"grad_norm": 11.52369059409131,
|
| 1350 |
+
"learning_rate": 2.9453681710213776e-07,
|
| 1351 |
+
"logits/chosen": -0.26953125,
|
| 1352 |
+
"logits/rejected": -0.359375,
|
| 1353 |
+
"logps/chosen": -400.0,
|
| 1354 |
+
"logps/rejected": -348.0,
|
| 1355 |
+
"loss": 0.052,
|
| 1356 |
+
"rewards/accuracies": 0.987500011920929,
|
| 1357 |
+
"rewards/chosen": 2.046875,
|
| 1358 |
+
"rewards/margins": 5.90625,
|
| 1359 |
+
"rewards/rejected": -3.84375,
|
| 1360 |
+
"step": 880
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 1.4262820512820513,
|
| 1364 |
+
"grad_norm": 15.524036946329323,
|
| 1365 |
+
"learning_rate": 2.915676959619953e-07,
|
| 1366 |
+
"logits/chosen": -0.1923828125,
|
| 1367 |
+
"logits/rejected": -0.283203125,
|
| 1368 |
+
"logps/chosen": -364.0,
|
| 1369 |
+
"logps/rejected": -358.0,
|
| 1370 |
+
"loss": 0.0819,
|
| 1371 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 1372 |
+
"rewards/chosen": 2.125,
|
| 1373 |
+
"rewards/margins": 6.1875,
|
| 1374 |
+
"rewards/rejected": -4.0625,
|
| 1375 |
+
"step": 890
|
| 1376 |
+
},
|
| 1377 |
+
{
|
| 1378 |
+
"epoch": 1.4423076923076923,
|
| 1379 |
+
"grad_norm": 56.41338438001929,
|
| 1380 |
+
"learning_rate": 2.885985748218527e-07,
|
| 1381 |
+
"logits/chosen": -0.2314453125,
|
| 1382 |
+
"logits/rejected": -0.427734375,
|
| 1383 |
+
"logps/chosen": -374.0,
|
| 1384 |
+
"logps/rejected": -352.0,
|
| 1385 |
+
"loss": 0.1086,
|
| 1386 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 1387 |
+
"rewards/chosen": 1.546875,
|
| 1388 |
+
"rewards/margins": 5.65625,
|
| 1389 |
+
"rewards/rejected": -4.125,
|
| 1390 |
+
"step": 900
|
| 1391 |
+
},
|
| 1392 |
+
{
|
| 1393 |
+
"epoch": 1.4583333333333333,
|
| 1394 |
+
"grad_norm": 35.802404269657636,
|
| 1395 |
+
"learning_rate": 2.856294536817102e-07,
|
| 1396 |
+
"logits/chosen": -0.169921875,
|
| 1397 |
+
"logits/rejected": -0.296875,
|
| 1398 |
+
"logps/chosen": -366.0,
|
| 1399 |
+
"logps/rejected": -352.0,
|
| 1400 |
+
"loss": 0.0569,
|
| 1401 |
+
"rewards/accuracies": 0.949999988079071,
|
| 1402 |
+
"rewards/chosen": 2.09375,
|
| 1403 |
+
"rewards/margins": 5.8125,
|
| 1404 |
+
"rewards/rejected": -3.71875,
|
| 1405 |
+
"step": 910
|
| 1406 |
+
},
|
| 1407 |
+
{
|
| 1408 |
+
"epoch": 1.4743589743589745,
|
| 1409 |
+
"grad_norm": 3.0990206807529574,
|
| 1410 |
+
"learning_rate": 2.8266033254156767e-07,
|
| 1411 |
+
"logits/chosen": -0.310546875,
|
| 1412 |
+
"logits/rejected": -0.16015625,
|
| 1413 |
+
"logps/chosen": -364.0,
|
| 1414 |
+
"logps/rejected": -358.0,
|
| 1415 |
+
"loss": 0.0532,
|
| 1416 |
+
"rewards/accuracies": 0.987500011920929,
|
| 1417 |
+
"rewards/chosen": 2.359375,
|
| 1418 |
+
"rewards/margins": 5.75,
|
| 1419 |
+
"rewards/rejected": -3.390625,
|
| 1420 |
+
"step": 920
|
| 1421 |
+
},
|
| 1422 |
+
{
|
| 1423 |
+
"epoch": 1.4903846153846154,
|
| 1424 |
+
"grad_norm": 6.630507067800798,
|
| 1425 |
+
"learning_rate": 2.7969121140142515e-07,
|
| 1426 |
+
"logits/chosen": -0.2099609375,
|
| 1427 |
+
"logits/rejected": -0.384765625,
|
| 1428 |
+
"logps/chosen": -352.0,
|
| 1429 |
+
"logps/rejected": -344.0,
|
| 1430 |
+
"loss": 0.1117,
|
| 1431 |
+
"rewards/accuracies": 0.949999988079071,
|
| 1432 |
+
"rewards/chosen": 2.78125,
|
| 1433 |
+
"rewards/margins": 5.9375,
|
| 1434 |
+
"rewards/rejected": -3.140625,
|
| 1435 |
+
"step": 930
|
| 1436 |
+
},
|
| 1437 |
+
{
|
| 1438 |
+
"epoch": 1.5064102564102564,
|
| 1439 |
+
"grad_norm": 3.9527473386787992,
|
| 1440 |
+
"learning_rate": 2.7672209026128263e-07,
|
| 1441 |
+
"logits/chosen": -0.150390625,
|
| 1442 |
+
"logits/rejected": -0.353515625,
|
| 1443 |
+
"logps/chosen": -364.0,
|
| 1444 |
+
"logps/rejected": -326.0,
|
| 1445 |
+
"loss": 0.1115,
|
| 1446 |
+
"rewards/accuracies": 0.949999988079071,
|
| 1447 |
+
"rewards/chosen": 2.71875,
|
| 1448 |
+
"rewards/margins": 5.375,
|
| 1449 |
+
"rewards/rejected": -2.65625,
|
| 1450 |
+
"step": 940
|
| 1451 |
+
},
|
| 1452 |
+
{
|
| 1453 |
+
"epoch": 1.5224358974358974,
|
| 1454 |
+
"grad_norm": 21.74505432942635,
|
| 1455 |
+
"learning_rate": 2.7375296912114016e-07,
|
| 1456 |
+
"logits/chosen": -0.224609375,
|
| 1457 |
+
"logits/rejected": -0.3203125,
|
| 1458 |
+
"logps/chosen": -388.0,
|
| 1459 |
+
"logps/rejected": -364.0,
|
| 1460 |
+
"loss": 0.0695,
|
| 1461 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 1462 |
+
"rewards/chosen": 2.765625,
|
| 1463 |
+
"rewards/margins": 5.5,
|
| 1464 |
+
"rewards/rejected": -2.75,
|
| 1465 |
+
"step": 950
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 1.5384615384615383,
|
| 1469 |
+
"grad_norm": 34.5728416770996,
|
| 1470 |
+
"learning_rate": 2.7078384798099764e-07,
|
| 1471 |
+
"logits/chosen": -0.2197265625,
|
| 1472 |
+
"logits/rejected": -0.412109375,
|
| 1473 |
+
"logps/chosen": -356.0,
|
| 1474 |
+
"logps/rejected": -328.0,
|
| 1475 |
+
"loss": 0.0797,
|
| 1476 |
+
"rewards/accuracies": 0.987500011920929,
|
| 1477 |
+
"rewards/chosen": 2.9375,
|
| 1478 |
+
"rewards/margins": 6.40625,
|
| 1479 |
+
"rewards/rejected": -3.46875,
|
| 1480 |
+
"step": 960
|
| 1481 |
+
},
|
| 1482 |
+
{
|
| 1483 |
+
"epoch": 1.5544871794871795,
|
| 1484 |
+
"grad_norm": 10.226160428385189,
|
| 1485 |
+
"learning_rate": 2.678147268408551e-07,
|
| 1486 |
+
"logits/chosen": -0.1787109375,
|
| 1487 |
+
"logits/rejected": -0.4140625,
|
| 1488 |
+
"logps/chosen": -372.0,
|
| 1489 |
+
"logps/rejected": -330.0,
|
| 1490 |
+
"loss": 0.0415,
|
| 1491 |
+
"rewards/accuracies": 0.987500011920929,
|
| 1492 |
+
"rewards/chosen": 2.40625,
|
| 1493 |
+
"rewards/margins": 6.40625,
|
| 1494 |
+
"rewards/rejected": -3.984375,
|
| 1495 |
+
"step": 970
|
| 1496 |
+
},
|
| 1497 |
+
{
|
| 1498 |
+
"epoch": 1.5705128205128205,
|
| 1499 |
+
"grad_norm": 2.887418080486524,
|
| 1500 |
+
"learning_rate": 2.648456057007126e-07,
|
| 1501 |
+
"logits/chosen": -0.1357421875,
|
| 1502 |
+
"logits/rejected": -0.318359375,
|
| 1503 |
+
"logps/chosen": -388.0,
|
| 1504 |
+
"logps/rejected": -320.0,
|
| 1505 |
+
"loss": 0.0622,
|
| 1506 |
+
"rewards/accuracies": 0.987500011920929,
|
| 1507 |
+
"rewards/chosen": 2.65625,
|
| 1508 |
+
"rewards/margins": 6.40625,
|
| 1509 |
+
"rewards/rejected": -3.734375,
|
| 1510 |
+
"step": 980
|
| 1511 |
+
},
|
| 1512 |
+
{
|
| 1513 |
+
"epoch": 1.5865384615384617,
|
| 1514 |
+
"grad_norm": 25.20155218665509,
|
| 1515 |
+
"learning_rate": 2.6187648456057e-07,
|
| 1516 |
+
"logits/chosen": -0.1044921875,
|
| 1517 |
+
"logits/rejected": -0.37890625,
|
| 1518 |
+
"logps/chosen": -330.0,
|
| 1519 |
+
"logps/rejected": -326.0,
|
| 1520 |
+
"loss": 0.0672,
|
| 1521 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 1522 |
+
"rewards/chosen": 2.921875,
|
| 1523 |
+
"rewards/margins": 6.75,
|
| 1524 |
+
"rewards/rejected": -3.828125,
|
| 1525 |
+
"step": 990
|
| 1526 |
+
},
|
| 1527 |
+
{
|
| 1528 |
+
"epoch": 1.6025641025641026,
|
| 1529 |
+
"grad_norm": 2.9326020885487107,
|
| 1530 |
+
"learning_rate": 2.589073634204275e-07,
|
| 1531 |
+
"logits/chosen": -0.244140625,
|
| 1532 |
+
"logits/rejected": -0.4609375,
|
| 1533 |
+
"logps/chosen": -370.0,
|
| 1534 |
+
"logps/rejected": -336.0,
|
| 1535 |
+
"loss": 0.0686,
|
| 1536 |
+
"rewards/accuracies": 0.949999988079071,
|
| 1537 |
+
"rewards/chosen": 2.4375,
|
| 1538 |
+
"rewards/margins": 6.375,
|
| 1539 |
+
"rewards/rejected": -3.9375,
|
| 1540 |
+
"step": 1000
|
| 1541 |
+
},
|
| 1542 |
+
{
|
| 1543 |
+
"epoch": 1.6185897435897436,
|
| 1544 |
+
"grad_norm": 2.6087260544675948,
|
| 1545 |
+
"learning_rate": 2.5593824228028503e-07,
|
| 1546 |
+
"logits/chosen": -0.330078125,
|
| 1547 |
+
"logits/rejected": -0.59765625,
|
| 1548 |
+
"logps/chosen": -366.0,
|
| 1549 |
+
"logps/rejected": -384.0,
|
| 1550 |
+
"loss": 0.0506,
|
| 1551 |
+
"rewards/accuracies": 0.987500011920929,
|
| 1552 |
+
"rewards/chosen": 2.390625,
|
| 1553 |
+
"rewards/margins": 6.4375,
|
| 1554 |
+
"rewards/rejected": -4.0625,
|
| 1555 |
+
"step": 1010
|
| 1556 |
+
},
|
| 1557 |
+
{
|
| 1558 |
+
"epoch": 1.6346153846153846,
|
| 1559 |
+
"grad_norm": 46.202092572333505,
|
| 1560 |
+
"learning_rate": 2.529691211401425e-07,
|
| 1561 |
+
"logits/chosen": -0.10791015625,
|
| 1562 |
+
"logits/rejected": -0.40234375,
|
| 1563 |
+
"logps/chosen": -370.0,
|
| 1564 |
+
"logps/rejected": -368.0,
|
| 1565 |
+
"loss": 0.1479,
|
| 1566 |
+
"rewards/accuracies": 0.9375,
|
| 1567 |
+
"rewards/chosen": 1.9609375,
|
| 1568 |
+
"rewards/margins": 5.6875,
|
| 1569 |
+
"rewards/rejected": -3.71875,
|
| 1570 |
+
"step": 1020
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 1.6506410256410255,
|
| 1574 |
+
"grad_norm": 11.039931814429327,
|
| 1575 |
+
"learning_rate": 2.5e-07,
|
| 1576 |
+
"logits/chosen": -0.1875,
|
| 1577 |
+
"logits/rejected": -0.357421875,
|
| 1578 |
+
"logps/chosen": -398.0,
|
| 1579 |
+
"logps/rejected": -348.0,
|
| 1580 |
+
"loss": 0.0695,
|
| 1581 |
+
"rewards/accuracies": 0.987500011920929,
|
| 1582 |
+
"rewards/chosen": 1.6875,
|
| 1583 |
+
"rewards/margins": 5.59375,
|
| 1584 |
+
"rewards/rejected": -3.90625,
|
| 1585 |
+
"step": 1030
|
| 1586 |
+
},
|
| 1587 |
+
{
|
| 1588 |
+
"epoch": 1.6666666666666665,
|
| 1589 |
+
"grad_norm": 19.29266338720831,
|
| 1590 |
+
"learning_rate": 2.4703087885985747e-07,
|
| 1591 |
+
"logits/chosen": -0.2353515625,
|
| 1592 |
+
"logits/rejected": -0.4609375,
|
| 1593 |
+
"logps/chosen": -364.0,
|
| 1594 |
+
"logps/rejected": -350.0,
|
| 1595 |
+
"loss": 0.0767,
|
| 1596 |
+
"rewards/accuracies": 0.987500011920929,
|
| 1597 |
+
"rewards/chosen": 1.5078125,
|
| 1598 |
+
"rewards/margins": 6.15625,
|
| 1599 |
+
"rewards/rejected": -4.65625,
|
| 1600 |
+
"step": 1040
|
| 1601 |
+
},
|
| 1602 |
+
{
|
| 1603 |
+
"epoch": 1.6826923076923077,
|
| 1604 |
+
"grad_norm": 42.04473353572841,
|
| 1605 |
+
"learning_rate": 2.4406175771971495e-07,
|
| 1606 |
+
"logits/chosen": -0.375,
|
| 1607 |
+
"logits/rejected": -0.515625,
|
| 1608 |
+
"logps/chosen": -374.0,
|
| 1609 |
+
"logps/rejected": -356.0,
|
| 1610 |
+
"loss": 0.0784,
|
| 1611 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 1612 |
+
"rewards/chosen": 1.8046875,
|
| 1613 |
+
"rewards/margins": 5.84375,
|
| 1614 |
+
"rewards/rejected": -4.0625,
|
| 1615 |
+
"step": 1050
|
| 1616 |
+
},
|
| 1617 |
+
{
|
| 1618 |
+
"epoch": 1.6987179487179487,
|
| 1619 |
+
"grad_norm": 48.05824905529893,
|
| 1620 |
+
"learning_rate": 2.410926365795724e-07,
|
| 1621 |
+
"logits/chosen": -0.2109375,
|
| 1622 |
+
"logits/rejected": -0.353515625,
|
| 1623 |
+
"logps/chosen": -380.0,
|
| 1624 |
+
"logps/rejected": -342.0,
|
| 1625 |
+
"loss": 0.0879,
|
| 1626 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 1627 |
+
"rewards/chosen": 1.7109375,
|
| 1628 |
+
"rewards/margins": 5.40625,
|
| 1629 |
+
"rewards/rejected": -3.703125,
|
| 1630 |
+
"step": 1060
|
| 1631 |
+
},
|
| 1632 |
+
{
|
| 1633 |
+
"epoch": 1.7147435897435899,
|
| 1634 |
+
"grad_norm": 9.491221178753944,
|
| 1635 |
+
"learning_rate": 2.381235154394299e-07,
|
| 1636 |
+
"logits/chosen": -0.2138671875,
|
| 1637 |
+
"logits/rejected": -0.4375,
|
| 1638 |
+
"logps/chosen": -372.0,
|
| 1639 |
+
"logps/rejected": -358.0,
|
| 1640 |
+
"loss": 0.0497,
|
| 1641 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 1642 |
+
"rewards/chosen": 2.296875,
|
| 1643 |
+
"rewards/margins": 6.4375,
|
| 1644 |
+
"rewards/rejected": -4.125,
|
| 1645 |
+
"step": 1070
|
| 1646 |
+
},
|
| 1647 |
+
{
|
| 1648 |
+
"epoch": 1.7307692307692308,
|
| 1649 |
+
"grad_norm": 40.31044273857979,
|
| 1650 |
+
"learning_rate": 2.351543942992874e-07,
|
| 1651 |
+
"logits/chosen": -0.248046875,
|
| 1652 |
+
"logits/rejected": -0.51953125,
|
| 1653 |
+
"logps/chosen": -360.0,
|
| 1654 |
+
"logps/rejected": -350.0,
|
| 1655 |
+
"loss": 0.1188,
|
| 1656 |
+
"rewards/accuracies": 0.949999988079071,
|
| 1657 |
+
"rewards/chosen": 2.515625,
|
| 1658 |
+
"rewards/margins": 6.34375,
|
| 1659 |
+
"rewards/rejected": -3.8125,
|
| 1660 |
+
"step": 1080
|
| 1661 |
+
},
|
| 1662 |
+
{
|
| 1663 |
+
"epoch": 1.7467948717948718,
|
| 1664 |
+
"grad_norm": 1.4266357038353779,
|
| 1665 |
+
"learning_rate": 2.3218527315914489e-07,
|
| 1666 |
+
"logits/chosen": -0.287109375,
|
| 1667 |
+
"logits/rejected": -0.283203125,
|
| 1668 |
+
"logps/chosen": -356.0,
|
| 1669 |
+
"logps/rejected": -328.0,
|
| 1670 |
+
"loss": 0.1075,
|
| 1671 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 1672 |
+
"rewards/chosen": 2.265625,
|
| 1673 |
+
"rewards/margins": 5.59375,
|
| 1674 |
+
"rewards/rejected": -3.328125,
|
| 1675 |
+
"step": 1090
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 1.7628205128205128,
|
| 1679 |
+
"grad_norm": 21.16673032403483,
|
| 1680 |
+
"learning_rate": 2.2921615201900234e-07,
|
| 1681 |
+
"logits/chosen": -0.1572265625,
|
| 1682 |
+
"logits/rejected": -0.408203125,
|
| 1683 |
+
"logps/chosen": -392.0,
|
| 1684 |
+
"logps/rejected": -322.0,
|
| 1685 |
+
"loss": 0.0501,
|
| 1686 |
+
"rewards/accuracies": 1.0,
|
| 1687 |
+
"rewards/chosen": 2.046875,
|
| 1688 |
+
"rewards/margins": 6.4375,
|
| 1689 |
+
"rewards/rejected": -4.40625,
|
| 1690 |
+
"step": 1100
|
| 1691 |
+
},
|
| 1692 |
+
{
|
| 1693 |
+
"epoch": 1.7788461538461537,
|
| 1694 |
+
"grad_norm": 32.08142550351567,
|
| 1695 |
+
"learning_rate": 2.2624703087885984e-07,
|
| 1696 |
+
"logits/chosen": -0.17578125,
|
| 1697 |
+
"logits/rejected": -0.373046875,
|
| 1698 |
+
"logps/chosen": -356.0,
|
| 1699 |
+
"logps/rejected": -348.0,
|
| 1700 |
+
"loss": 0.0622,
|
| 1701 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 1702 |
+
"rewards/chosen": 2.0625,
|
| 1703 |
+
"rewards/margins": 6.3125,
|
| 1704 |
+
"rewards/rejected": -4.25,
|
| 1705 |
+
"step": 1110
|
| 1706 |
+
},
|
| 1707 |
+
{
|
| 1708 |
+
"epoch": 1.7948717948717947,
|
| 1709 |
+
"grad_norm": 10.94873802284618,
|
| 1710 |
+
"learning_rate": 2.2327790973871732e-07,
|
| 1711 |
+
"logits/chosen": -0.296875,
|
| 1712 |
+
"logits/rejected": -0.474609375,
|
| 1713 |
+
"logps/chosen": -372.0,
|
| 1714 |
+
"logps/rejected": -368.0,
|
| 1715 |
+
"loss": 0.0631,
|
| 1716 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 1717 |
+
"rewards/chosen": 2.015625,
|
| 1718 |
+
"rewards/margins": 6.15625,
|
| 1719 |
+
"rewards/rejected": -4.125,
|
| 1720 |
+
"step": 1120
|
| 1721 |
+
},
|
| 1722 |
+
{
|
| 1723 |
+
"epoch": 1.810897435897436,
|
| 1724 |
+
"grad_norm": 4.164714441310206,
|
| 1725 |
+
"learning_rate": 2.2030878859857483e-07,
|
| 1726 |
+
"logits/chosen": -0.251953125,
|
| 1727 |
+
"logits/rejected": -0.330078125,
|
| 1728 |
+
"logps/chosen": -362.0,
|
| 1729 |
+
"logps/rejected": -346.0,
|
| 1730 |
+
"loss": 0.0571,
|
| 1731 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 1732 |
+
"rewards/chosen": 2.015625,
|
| 1733 |
+
"rewards/margins": 6.21875,
|
| 1734 |
+
"rewards/rejected": -4.1875,
|
| 1735 |
+
"step": 1130
|
| 1736 |
+
},
|
| 1737 |
+
{
|
| 1738 |
+
"epoch": 1.8269230769230769,
|
| 1739 |
+
"grad_norm": 6.828795713228818,
|
| 1740 |
+
"learning_rate": 2.173396674584323e-07,
|
| 1741 |
+
"logits/chosen": -0.234375,
|
| 1742 |
+
"logits/rejected": -0.33984375,
|
| 1743 |
+
"logps/chosen": -376.0,
|
| 1744 |
+
"logps/rejected": -352.0,
|
| 1745 |
+
"loss": 0.1227,
|
| 1746 |
+
"rewards/accuracies": 0.949999988079071,
|
| 1747 |
+
"rewards/chosen": 2.234375,
|
| 1748 |
+
"rewards/margins": 5.5,
|
| 1749 |
+
"rewards/rejected": -3.234375,
|
| 1750 |
+
"step": 1140
|
| 1751 |
+
},
|
| 1752 |
+
{
|
| 1753 |
+
"epoch": 1.842948717948718,
|
| 1754 |
+
"grad_norm": 34.413780149566165,
|
| 1755 |
+
"learning_rate": 2.1437054631828976e-07,
|
| 1756 |
+
"logits/chosen": -0.224609375,
|
| 1757 |
+
"logits/rejected": -0.296875,
|
| 1758 |
+
"logps/chosen": -390.0,
|
| 1759 |
+
"logps/rejected": -346.0,
|
| 1760 |
+
"loss": 0.0492,
|
| 1761 |
+
"rewards/accuracies": 1.0,
|
| 1762 |
+
"rewards/chosen": 1.640625,
|
| 1763 |
+
"rewards/margins": 6.1875,
|
| 1764 |
+
"rewards/rejected": -4.5625,
|
| 1765 |
+
"step": 1150
|
| 1766 |
+
},
|
| 1767 |
+
{
|
| 1768 |
+
"epoch": 1.858974358974359,
|
| 1769 |
+
"grad_norm": 3.332529440650818,
|
| 1770 |
+
"learning_rate": 2.1140142517814726e-07,
|
| 1771 |
+
"logits/chosen": -0.1669921875,
|
| 1772 |
+
"logits/rejected": -0.376953125,
|
| 1773 |
+
"logps/chosen": -352.0,
|
| 1774 |
+
"logps/rejected": -370.0,
|
| 1775 |
+
"loss": 0.0768,
|
| 1776 |
+
"rewards/accuracies": 0.987500011920929,
|
| 1777 |
+
"rewards/chosen": 1.875,
|
| 1778 |
+
"rewards/margins": 6.34375,
|
| 1779 |
+
"rewards/rejected": -4.46875,
|
| 1780 |
+
"step": 1160
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 1.875,
|
| 1784 |
+
"grad_norm": 8.41589370404434,
|
| 1785 |
+
"learning_rate": 2.0843230403800474e-07,
|
| 1786 |
+
"logits/chosen": -0.349609375,
|
| 1787 |
+
"logits/rejected": -0.31640625,
|
| 1788 |
+
"logps/chosen": -374.0,
|
| 1789 |
+
"logps/rejected": -350.0,
|
| 1790 |
+
"loss": 0.0811,
|
| 1791 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 1792 |
+
"rewards/chosen": 1.7578125,
|
| 1793 |
+
"rewards/margins": 5.5,
|
| 1794 |
+
"rewards/rejected": -3.75,
|
| 1795 |
+
"step": 1170
|
| 1796 |
+
},
|
| 1797 |
+
{
|
| 1798 |
+
"epoch": 1.891025641025641,
|
| 1799 |
+
"grad_norm": 40.47797504678685,
|
| 1800 |
+
"learning_rate": 2.0546318289786222e-07,
|
| 1801 |
+
"logits/chosen": -0.2041015625,
|
| 1802 |
+
"logits/rejected": -0.40234375,
|
| 1803 |
+
"logps/chosen": -354.0,
|
| 1804 |
+
"logps/rejected": -366.0,
|
| 1805 |
+
"loss": 0.1088,
|
| 1806 |
+
"rewards/accuracies": 0.9375,
|
| 1807 |
+
"rewards/chosen": 1.5703125,
|
| 1808 |
+
"rewards/margins": 5.53125,
|
| 1809 |
+
"rewards/rejected": -3.953125,
|
| 1810 |
+
"step": 1180
|
| 1811 |
+
},
|
| 1812 |
+
{
|
| 1813 |
+
"epoch": 1.907051282051282,
|
| 1814 |
+
"grad_norm": 8.055421977487784,
|
| 1815 |
+
"learning_rate": 2.0249406175771972e-07,
|
| 1816 |
+
"logits/chosen": -0.298828125,
|
| 1817 |
+
"logits/rejected": -0.482421875,
|
| 1818 |
+
"logps/chosen": -380.0,
|
| 1819 |
+
"logps/rejected": -354.0,
|
| 1820 |
+
"loss": 0.0963,
|
| 1821 |
+
"rewards/accuracies": 0.987500011920929,
|
| 1822 |
+
"rewards/chosen": 2.046875,
|
| 1823 |
+
"rewards/margins": 6.0625,
|
| 1824 |
+
"rewards/rejected": -4.0,
|
| 1825 |
+
"step": 1190
|
| 1826 |
+
},
|
| 1827 |
+
{
|
| 1828 |
+
"epoch": 1.9230769230769231,
|
| 1829 |
+
"grad_norm": 1.0123224706730725,
|
| 1830 |
+
"learning_rate": 1.9952494061757718e-07,
|
| 1831 |
+
"logits/chosen": -0.259765625,
|
| 1832 |
+
"logits/rejected": -0.10888671875,
|
| 1833 |
+
"logps/chosen": -348.0,
|
| 1834 |
+
"logps/rejected": -356.0,
|
| 1835 |
+
"loss": 0.0319,
|
| 1836 |
+
"rewards/accuracies": 0.987500011920929,
|
| 1837 |
+
"rewards/chosen": 2.71875,
|
| 1838 |
+
"rewards/margins": 7.03125,
|
| 1839 |
+
"rewards/rejected": -4.3125,
|
| 1840 |
+
"step": 1200
|
| 1841 |
+
},
|
| 1842 |
+
{
|
| 1843 |
+
"epoch": 1.939102564102564,
|
| 1844 |
+
"grad_norm": 44.45376114208627,
|
| 1845 |
+
"learning_rate": 1.9655581947743466e-07,
|
| 1846 |
+
"logits/chosen": -0.21875,
|
| 1847 |
+
"logits/rejected": -0.3671875,
|
| 1848 |
+
"logps/chosen": -372.0,
|
| 1849 |
+
"logps/rejected": -376.0,
|
| 1850 |
+
"loss": 0.0746,
|
| 1851 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 1852 |
+
"rewards/chosen": 1.7578125,
|
| 1853 |
+
"rewards/margins": 6.59375,
|
| 1854 |
+
"rewards/rejected": -4.8125,
|
| 1855 |
+
"step": 1210
|
| 1856 |
+
},
|
| 1857 |
+
{
|
| 1858 |
+
"epoch": 1.9551282051282053,
|
| 1859 |
+
"grad_norm": 15.603585643576269,
|
| 1860 |
+
"learning_rate": 1.9358669833729216e-07,
|
| 1861 |
+
"logits/chosen": -0.15234375,
|
| 1862 |
+
"logits/rejected": -0.373046875,
|
| 1863 |
+
"logps/chosen": -380.0,
|
| 1864 |
+
"logps/rejected": -352.0,
|
| 1865 |
+
"loss": 0.0608,
|
| 1866 |
+
"rewards/accuracies": 0.987500011920929,
|
| 1867 |
+
"rewards/chosen": 1.734375,
|
| 1868 |
+
"rewards/margins": 6.0625,
|
| 1869 |
+
"rewards/rejected": -4.3125,
|
| 1870 |
+
"step": 1220
|
| 1871 |
+
},
|
| 1872 |
+
{
|
| 1873 |
+
"epoch": 1.9711538461538463,
|
| 1874 |
+
"grad_norm": 13.268798051575008,
|
| 1875 |
+
"learning_rate": 1.9061757719714964e-07,
|
| 1876 |
+
"logits/chosen": -0.380859375,
|
| 1877 |
+
"logits/rejected": -0.45703125,
|
| 1878 |
+
"logps/chosen": -396.0,
|
| 1879 |
+
"logps/rejected": -362.0,
|
| 1880 |
+
"loss": 0.0417,
|
| 1881 |
+
"rewards/accuracies": 1.0,
|
| 1882 |
+
"rewards/chosen": 2.328125,
|
| 1883 |
+
"rewards/margins": 6.65625,
|
| 1884 |
+
"rewards/rejected": -4.3125,
|
| 1885 |
+
"step": 1230
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 1.9871794871794872,
|
| 1889 |
+
"grad_norm": 18.53274384798147,
|
| 1890 |
+
"learning_rate": 1.876484560570071e-07,
|
| 1891 |
+
"logits/chosen": -0.1357421875,
|
| 1892 |
+
"logits/rejected": -0.251953125,
|
| 1893 |
+
"logps/chosen": -390.0,
|
| 1894 |
+
"logps/rejected": -342.0,
|
| 1895 |
+
"loss": 0.0639,
|
| 1896 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 1897 |
+
"rewards/chosen": 2.53125,
|
| 1898 |
+
"rewards/margins": 6.78125,
|
| 1899 |
+
"rewards/rejected": -4.25,
|
| 1900 |
+
"step": 1240
|
| 1901 |
+
},
|
| 1902 |
+
{
|
| 1903 |
+
"epoch": 2.0,
|
| 1904 |
+
"eval_logits/chosen": -0.25,
|
| 1905 |
+
"eval_logits/rejected": -0.30078125,
|
| 1906 |
+
"eval_logps/chosen": -362.0,
|
| 1907 |
+
"eval_logps/rejected": -358.0,
|
| 1908 |
+
"eval_loss": 0.2055511474609375,
|
| 1909 |
+
"eval_rewards/accuracies": 0.9038461446762085,
|
| 1910 |
+
"eval_rewards/chosen": 1.859375,
|
| 1911 |
+
"eval_rewards/margins": 5.4375,
|
| 1912 |
+
"eval_rewards/rejected": -3.5625,
|
| 1913 |
+
"eval_runtime": 29.021,
|
| 1914 |
+
"eval_samples_per_second": 6.892,
|
| 1915 |
+
"eval_steps_per_second": 0.448,
|
| 1916 |
+
"step": 1248
|
| 1917 |
+
},
|
| 1918 |
+
{
|
| 1919 |
+
"epoch": 2.003205128205128,
|
| 1920 |
+
"grad_norm": 2.3679308142140885,
|
| 1921 |
+
"learning_rate": 1.846793349168646e-07,
|
| 1922 |
+
"logits/chosen": -0.341796875,
|
| 1923 |
+
"logits/rejected": -0.46484375,
|
| 1924 |
+
"logps/chosen": -366.0,
|
| 1925 |
+
"logps/rejected": -352.0,
|
| 1926 |
+
"loss": 0.0315,
|
| 1927 |
+
"rewards/accuracies": 0.987500011920929,
|
| 1928 |
+
"rewards/chosen": 2.609375,
|
| 1929 |
+
"rewards/margins": 6.9375,
|
| 1930 |
+
"rewards/rejected": -4.3125,
|
| 1931 |
+
"step": 1250
|
| 1932 |
+
},
|
| 1933 |
+
{
|
| 1934 |
+
"epoch": 2.019230769230769,
|
| 1935 |
+
"grad_norm": 2.590719502204235,
|
| 1936 |
+
"learning_rate": 1.8171021377672207e-07,
|
| 1937 |
+
"logits/chosen": -0.25,
|
| 1938 |
+
"logits/rejected": -0.404296875,
|
| 1939 |
+
"logps/chosen": -378.0,
|
| 1940 |
+
"logps/rejected": -364.0,
|
| 1941 |
+
"loss": 0.0081,
|
| 1942 |
+
"rewards/accuracies": 1.0,
|
| 1943 |
+
"rewards/chosen": 2.015625,
|
| 1944 |
+
"rewards/margins": 6.90625,
|
| 1945 |
+
"rewards/rejected": -4.875,
|
| 1946 |
+
"step": 1260
|
| 1947 |
+
},
|
| 1948 |
+
{
|
| 1949 |
+
"epoch": 2.03525641025641,
|
| 1950 |
+
"grad_norm": 22.175507349451703,
|
| 1951 |
+
"learning_rate": 1.7874109263657958e-07,
|
| 1952 |
+
"logits/chosen": -0.2275390625,
|
| 1953 |
+
"logits/rejected": -0.37890625,
|
| 1954 |
+
"logps/chosen": -364.0,
|
| 1955 |
+
"logps/rejected": -384.0,
|
| 1956 |
+
"loss": 0.0156,
|
| 1957 |
+
"rewards/accuracies": 0.987500011920929,
|
| 1958 |
+
"rewards/chosen": 2.0625,
|
| 1959 |
+
"rewards/margins": 7.375,
|
| 1960 |
+
"rewards/rejected": -5.3125,
|
| 1961 |
+
"step": 1270
|
| 1962 |
+
},
|
| 1963 |
+
{
|
| 1964 |
+
"epoch": 2.051282051282051,
|
| 1965 |
+
"grad_norm": 4.568251724068892,
|
| 1966 |
+
"learning_rate": 1.7577197149643706e-07,
|
| 1967 |
+
"logits/chosen": -0.25,
|
| 1968 |
+
"logits/rejected": -0.482421875,
|
| 1969 |
+
"logps/chosen": -380.0,
|
| 1970 |
+
"logps/rejected": -370.0,
|
| 1971 |
+
"loss": 0.0145,
|
| 1972 |
+
"rewards/accuracies": 1.0,
|
| 1973 |
+
"rewards/chosen": 1.953125,
|
| 1974 |
+
"rewards/margins": 7.46875,
|
| 1975 |
+
"rewards/rejected": -5.53125,
|
| 1976 |
+
"step": 1280
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 2.0673076923076925,
|
| 1980 |
+
"grad_norm": 4.904685892165479,
|
| 1981 |
+
"learning_rate": 1.728028503562945e-07,
|
| 1982 |
+
"logits/chosen": -0.22265625,
|
| 1983 |
+
"logits/rejected": -0.466796875,
|
| 1984 |
+
"logps/chosen": -376.0,
|
| 1985 |
+
"logps/rejected": -366.0,
|
| 1986 |
+
"loss": 0.0088,
|
| 1987 |
+
"rewards/accuracies": 1.0,
|
| 1988 |
+
"rewards/chosen": 2.515625,
|
| 1989 |
+
"rewards/margins": 7.3125,
|
| 1990 |
+
"rewards/rejected": -4.78125,
|
| 1991 |
+
"step": 1290
|
| 1992 |
+
},
|
| 1993 |
+
{
|
| 1994 |
+
"epoch": 2.0833333333333335,
|
| 1995 |
+
"grad_norm": 0.6187256370987634,
|
| 1996 |
+
"learning_rate": 1.6983372921615202e-07,
|
| 1997 |
+
"logits/chosen": -0.2138671875,
|
| 1998 |
+
"logits/rejected": -0.458984375,
|
| 1999 |
+
"logps/chosen": -362.0,
|
| 2000 |
+
"logps/rejected": -370.0,
|
| 2001 |
+
"loss": 0.0123,
|
| 2002 |
+
"rewards/accuracies": 1.0,
|
| 2003 |
+
"rewards/chosen": 2.265625,
|
| 2004 |
+
"rewards/margins": 7.625,
|
| 2005 |
+
"rewards/rejected": -5.375,
|
| 2006 |
+
"step": 1300
|
| 2007 |
+
},
|
| 2008 |
+
{
|
| 2009 |
+
"epoch": 2.0993589743589745,
|
| 2010 |
+
"grad_norm": 0.7331904803685036,
|
| 2011 |
+
"learning_rate": 1.668646080760095e-07,
|
| 2012 |
+
"logits/chosen": -0.15234375,
|
| 2013 |
+
"logits/rejected": -0.40625,
|
| 2014 |
+
"logps/chosen": -354.0,
|
| 2015 |
+
"logps/rejected": -356.0,
|
| 2016 |
+
"loss": 0.0061,
|
| 2017 |
+
"rewards/accuracies": 1.0,
|
| 2018 |
+
"rewards/chosen": 2.59375,
|
| 2019 |
+
"rewards/margins": 7.65625,
|
| 2020 |
+
"rewards/rejected": -5.0625,
|
| 2021 |
+
"step": 1310
|
| 2022 |
+
},
|
| 2023 |
+
{
|
| 2024 |
+
"epoch": 2.1153846153846154,
|
| 2025 |
+
"grad_norm": 2.0296944258203036,
|
| 2026 |
+
"learning_rate": 1.6389548693586697e-07,
|
| 2027 |
+
"logits/chosen": -0.2412109375,
|
| 2028 |
+
"logits/rejected": -0.1748046875,
|
| 2029 |
+
"logps/chosen": -348.0,
|
| 2030 |
+
"logps/rejected": -348.0,
|
| 2031 |
+
"loss": 0.0231,
|
| 2032 |
+
"rewards/accuracies": 0.987500011920929,
|
| 2033 |
+
"rewards/chosen": 2.484375,
|
| 2034 |
+
"rewards/margins": 7.125,
|
| 2035 |
+
"rewards/rejected": -4.625,
|
| 2036 |
+
"step": 1320
|
| 2037 |
+
},
|
| 2038 |
+
{
|
| 2039 |
+
"epoch": 2.1314102564102564,
|
| 2040 |
+
"grad_norm": 23.71632295557207,
|
| 2041 |
+
"learning_rate": 1.6092636579572448e-07,
|
| 2042 |
+
"logits/chosen": -0.2275390625,
|
| 2043 |
+
"logits/rejected": -0.3125,
|
| 2044 |
+
"logps/chosen": -358.0,
|
| 2045 |
+
"logps/rejected": -342.0,
|
| 2046 |
+
"loss": 0.019,
|
| 2047 |
+
"rewards/accuracies": 1.0,
|
| 2048 |
+
"rewards/chosen": 2.78125,
|
| 2049 |
+
"rewards/margins": 7.90625,
|
| 2050 |
+
"rewards/rejected": -5.125,
|
| 2051 |
+
"step": 1330
|
| 2052 |
+
},
|
| 2053 |
+
{
|
| 2054 |
+
"epoch": 2.1474358974358974,
|
| 2055 |
+
"grad_norm": 4.284093803737362,
|
| 2056 |
+
"learning_rate": 1.5795724465558193e-07,
|
| 2057 |
+
"logits/chosen": -0.2255859375,
|
| 2058 |
+
"logits/rejected": -0.5546875,
|
| 2059 |
+
"logps/chosen": -364.0,
|
| 2060 |
+
"logps/rejected": -384.0,
|
| 2061 |
+
"loss": 0.0159,
|
| 2062 |
+
"rewards/accuracies": 1.0,
|
| 2063 |
+
"rewards/chosen": 2.484375,
|
| 2064 |
+
"rewards/margins": 7.4375,
|
| 2065 |
+
"rewards/rejected": -4.9375,
|
| 2066 |
+
"step": 1340
|
| 2067 |
+
},
|
| 2068 |
+
{
|
| 2069 |
+
"epoch": 2.1634615384615383,
|
| 2070 |
+
"grad_norm": 0.7862983762455222,
|
| 2071 |
+
"learning_rate": 1.549881235154394e-07,
|
| 2072 |
+
"logits/chosen": -0.1435546875,
|
| 2073 |
+
"logits/rejected": -0.337890625,
|
| 2074 |
+
"logps/chosen": -376.0,
|
| 2075 |
+
"logps/rejected": -354.0,
|
| 2076 |
+
"loss": 0.0114,
|
| 2077 |
+
"rewards/accuracies": 1.0,
|
| 2078 |
+
"rewards/chosen": 2.828125,
|
| 2079 |
+
"rewards/margins": 7.40625,
|
| 2080 |
+
"rewards/rejected": -4.59375,
|
| 2081 |
+
"step": 1350
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 2.1794871794871793,
|
| 2085 |
+
"grad_norm": 1.8750426567792726,
|
| 2086 |
+
"learning_rate": 1.520190023752969e-07,
|
| 2087 |
+
"logits/chosen": -0.259765625,
|
| 2088 |
+
"logits/rejected": -0.59765625,
|
| 2089 |
+
"logps/chosen": -364.0,
|
| 2090 |
+
"logps/rejected": -330.0,
|
| 2091 |
+
"loss": 0.0138,
|
| 2092 |
+
"rewards/accuracies": 1.0,
|
| 2093 |
+
"rewards/chosen": 3.015625,
|
| 2094 |
+
"rewards/margins": 8.1875,
|
| 2095 |
+
"rewards/rejected": -5.15625,
|
| 2096 |
+
"step": 1360
|
| 2097 |
+
},
|
| 2098 |
+
{
|
| 2099 |
+
"epoch": 2.1955128205128207,
|
| 2100 |
+
"grad_norm": 7.170950559094129,
|
| 2101 |
+
"learning_rate": 1.490498812351544e-07,
|
| 2102 |
+
"logits/chosen": -0.251953125,
|
| 2103 |
+
"logits/rejected": -0.36328125,
|
| 2104 |
+
"logps/chosen": -362.0,
|
| 2105 |
+
"logps/rejected": -356.0,
|
| 2106 |
+
"loss": 0.0323,
|
| 2107 |
+
"rewards/accuracies": 1.0,
|
| 2108 |
+
"rewards/chosen": 3.0,
|
| 2109 |
+
"rewards/margins": 8.1875,
|
| 2110 |
+
"rewards/rejected": -5.1875,
|
| 2111 |
+
"step": 1370
|
| 2112 |
+
},
|
| 2113 |
+
{
|
| 2114 |
+
"epoch": 2.2115384615384617,
|
| 2115 |
+
"grad_norm": 21.23713813309802,
|
| 2116 |
+
"learning_rate": 1.4608076009501184e-07,
|
| 2117 |
+
"logits/chosen": -0.2431640625,
|
| 2118 |
+
"logits/rejected": -0.2353515625,
|
| 2119 |
+
"logps/chosen": -366.0,
|
| 2120 |
+
"logps/rejected": -356.0,
|
| 2121 |
+
"loss": 0.024,
|
| 2122 |
+
"rewards/accuracies": 1.0,
|
| 2123 |
+
"rewards/chosen": 2.6875,
|
| 2124 |
+
"rewards/margins": 7.8125,
|
| 2125 |
+
"rewards/rejected": -5.125,
|
| 2126 |
+
"step": 1380
|
| 2127 |
+
},
|
| 2128 |
+
{
|
| 2129 |
+
"epoch": 2.2275641025641026,
|
| 2130 |
+
"grad_norm": 7.169471258104354,
|
| 2131 |
+
"learning_rate": 1.4311163895486935e-07,
|
| 2132 |
+
"logits/chosen": -0.18359375,
|
| 2133 |
+
"logits/rejected": -0.29296875,
|
| 2134 |
+
"logps/chosen": -348.0,
|
| 2135 |
+
"logps/rejected": -364.0,
|
| 2136 |
+
"loss": 0.0257,
|
| 2137 |
+
"rewards/accuracies": 1.0,
|
| 2138 |
+
"rewards/chosen": 2.90625,
|
| 2139 |
+
"rewards/margins": 8.1875,
|
| 2140 |
+
"rewards/rejected": -5.25,
|
| 2141 |
+
"step": 1390
|
| 2142 |
+
},
|
| 2143 |
+
{
|
| 2144 |
+
"epoch": 2.2435897435897436,
|
| 2145 |
+
"grad_norm": 4.891546899421327,
|
| 2146 |
+
"learning_rate": 1.4014251781472683e-07,
|
| 2147 |
+
"logits/chosen": -0.22265625,
|
| 2148 |
+
"logits/rejected": -0.55078125,
|
| 2149 |
+
"logps/chosen": -344.0,
|
| 2150 |
+
"logps/rejected": -374.0,
|
| 2151 |
+
"loss": 0.0165,
|
| 2152 |
+
"rewards/accuracies": 0.987500011920929,
|
| 2153 |
+
"rewards/chosen": 2.71875,
|
| 2154 |
+
"rewards/margins": 7.4375,
|
| 2155 |
+
"rewards/rejected": -4.71875,
|
| 2156 |
+
"step": 1400
|
| 2157 |
+
},
|
| 2158 |
+
{
|
| 2159 |
+
"epoch": 2.2596153846153846,
|
| 2160 |
+
"grad_norm": 14.0088561337579,
|
| 2161 |
+
"learning_rate": 1.3717339667458433e-07,
|
| 2162 |
+
"logits/chosen": -0.171875,
|
| 2163 |
+
"logits/rejected": -0.375,
|
| 2164 |
+
"logps/chosen": -336.0,
|
| 2165 |
+
"logps/rejected": -380.0,
|
| 2166 |
+
"loss": 0.023,
|
| 2167 |
+
"rewards/accuracies": 1.0,
|
| 2168 |
+
"rewards/chosen": 2.328125,
|
| 2169 |
+
"rewards/margins": 8.125,
|
| 2170 |
+
"rewards/rejected": -5.8125,
|
| 2171 |
+
"step": 1410
|
| 2172 |
+
},
|
| 2173 |
+
{
|
| 2174 |
+
"epoch": 2.2756410256410255,
|
| 2175 |
+
"grad_norm": 5.478498841262275,
|
| 2176 |
+
"learning_rate": 1.342042755344418e-07,
|
| 2177 |
+
"logits/chosen": -0.244140625,
|
| 2178 |
+
"logits/rejected": -0.60546875,
|
| 2179 |
+
"logps/chosen": -340.0,
|
| 2180 |
+
"logps/rejected": -368.0,
|
| 2181 |
+
"loss": 0.0101,
|
| 2182 |
+
"rewards/accuracies": 1.0,
|
| 2183 |
+
"rewards/chosen": 2.078125,
|
| 2184 |
+
"rewards/margins": 7.125,
|
| 2185 |
+
"rewards/rejected": -5.0625,
|
| 2186 |
+
"step": 1420
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 2.2916666666666665,
|
| 2190 |
+
"grad_norm": 4.08870815918569,
|
| 2191 |
+
"learning_rate": 1.3123515439429926e-07,
|
| 2192 |
+
"logits/chosen": -0.216796875,
|
| 2193 |
+
"logits/rejected": -0.5,
|
| 2194 |
+
"logps/chosen": -348.0,
|
| 2195 |
+
"logps/rejected": -380.0,
|
| 2196 |
+
"loss": 0.0094,
|
| 2197 |
+
"rewards/accuracies": 1.0,
|
| 2198 |
+
"rewards/chosen": 2.40625,
|
| 2199 |
+
"rewards/margins": 7.875,
|
| 2200 |
+
"rewards/rejected": -5.46875,
|
| 2201 |
+
"step": 1430
|
| 2202 |
+
},
|
| 2203 |
+
{
|
| 2204 |
+
"epoch": 2.3076923076923075,
|
| 2205 |
+
"grad_norm": 4.0828939935609885,
|
| 2206 |
+
"learning_rate": 1.2826603325415677e-07,
|
| 2207 |
+
"logits/chosen": -0.310546875,
|
| 2208 |
+
"logits/rejected": -0.3515625,
|
| 2209 |
+
"logps/chosen": -362.0,
|
| 2210 |
+
"logps/rejected": -372.0,
|
| 2211 |
+
"loss": 0.0115,
|
| 2212 |
+
"rewards/accuracies": 1.0,
|
| 2213 |
+
"rewards/chosen": 1.9296875,
|
| 2214 |
+
"rewards/margins": 7.90625,
|
| 2215 |
+
"rewards/rejected": -5.96875,
|
| 2216 |
+
"step": 1440
|
| 2217 |
+
},
|
| 2218 |
+
{
|
| 2219 |
+
"epoch": 2.323717948717949,
|
| 2220 |
+
"grad_norm": 2.2611774310220714,
|
| 2221 |
+
"learning_rate": 1.2529691211401425e-07,
|
| 2222 |
+
"logits/chosen": -0.2890625,
|
| 2223 |
+
"logits/rejected": -0.3671875,
|
| 2224 |
+
"logps/chosen": -350.0,
|
| 2225 |
+
"logps/rejected": -380.0,
|
| 2226 |
+
"loss": 0.0071,
|
| 2227 |
+
"rewards/accuracies": 1.0,
|
| 2228 |
+
"rewards/chosen": 2.359375,
|
| 2229 |
+
"rewards/margins": 8.1875,
|
| 2230 |
+
"rewards/rejected": -5.84375,
|
| 2231 |
+
"step": 1450
|
| 2232 |
+
},
|
| 2233 |
+
{
|
| 2234 |
+
"epoch": 2.33974358974359,
|
| 2235 |
+
"grad_norm": 35.51723177781465,
|
| 2236 |
+
"learning_rate": 1.2232779097387173e-07,
|
| 2237 |
+
"logits/chosen": -0.244140625,
|
| 2238 |
+
"logits/rejected": -0.36328125,
|
| 2239 |
+
"logps/chosen": -378.0,
|
| 2240 |
+
"logps/rejected": -368.0,
|
| 2241 |
+
"loss": 0.0773,
|
| 2242 |
+
"rewards/accuracies": 1.0,
|
| 2243 |
+
"rewards/chosen": 2.078125,
|
| 2244 |
+
"rewards/margins": 8.125,
|
| 2245 |
+
"rewards/rejected": -6.0,
|
| 2246 |
+
"step": 1460
|
| 2247 |
+
},
|
| 2248 |
+
{
|
| 2249 |
+
"epoch": 2.355769230769231,
|
| 2250 |
+
"grad_norm": 5.534009913713299,
|
| 2251 |
+
"learning_rate": 1.193586698337292e-07,
|
| 2252 |
+
"logits/chosen": -0.1767578125,
|
| 2253 |
+
"logits/rejected": -0.5390625,
|
| 2254 |
+
"logps/chosen": -368.0,
|
| 2255 |
+
"logps/rejected": -358.0,
|
| 2256 |
+
"loss": 0.012,
|
| 2257 |
+
"rewards/accuracies": 1.0,
|
| 2258 |
+
"rewards/chosen": 2.484375,
|
| 2259 |
+
"rewards/margins": 8.0625,
|
| 2260 |
+
"rewards/rejected": -5.59375,
|
| 2261 |
+
"step": 1470
|
| 2262 |
+
},
|
| 2263 |
+
{
|
| 2264 |
+
"epoch": 2.371794871794872,
|
| 2265 |
+
"grad_norm": 18.31189968205467,
|
| 2266 |
+
"learning_rate": 1.163895486935867e-07,
|
| 2267 |
+
"logits/chosen": -0.25390625,
|
| 2268 |
+
"logits/rejected": -0.353515625,
|
| 2269 |
+
"logps/chosen": -368.0,
|
| 2270 |
+
"logps/rejected": -356.0,
|
| 2271 |
+
"loss": 0.0359,
|
| 2272 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 2273 |
+
"rewards/chosen": 2.3125,
|
| 2274 |
+
"rewards/margins": 7.78125,
|
| 2275 |
+
"rewards/rejected": -5.46875,
|
| 2276 |
+
"step": 1480
|
| 2277 |
+
},
|
| 2278 |
+
{
|
| 2279 |
+
"epoch": 2.3878205128205128,
|
| 2280 |
+
"grad_norm": 14.190561755907082,
|
| 2281 |
+
"learning_rate": 1.1342042755344417e-07,
|
| 2282 |
+
"logits/chosen": -0.1484375,
|
| 2283 |
+
"logits/rejected": -0.35546875,
|
| 2284 |
+
"logps/chosen": -376.0,
|
| 2285 |
+
"logps/rejected": -370.0,
|
| 2286 |
+
"loss": 0.0065,
|
| 2287 |
+
"rewards/accuracies": 1.0,
|
| 2288 |
+
"rewards/chosen": 2.734375,
|
| 2289 |
+
"rewards/margins": 8.4375,
|
| 2290 |
+
"rewards/rejected": -5.6875,
|
| 2291 |
+
"step": 1490
|
| 2292 |
+
},
|
| 2293 |
+
{
|
| 2294 |
+
"epoch": 2.4038461538461537,
|
| 2295 |
+
"grad_norm": 57.38633989383038,
|
| 2296 |
+
"learning_rate": 1.1045130641330165e-07,
|
| 2297 |
+
"logits/chosen": -0.2177734375,
|
| 2298 |
+
"logits/rejected": -0.24609375,
|
| 2299 |
+
"logps/chosen": -356.0,
|
| 2300 |
+
"logps/rejected": -376.0,
|
| 2301 |
+
"loss": 0.0762,
|
| 2302 |
+
"rewards/accuracies": 0.949999988079071,
|
| 2303 |
+
"rewards/chosen": 1.25,
|
| 2304 |
+
"rewards/margins": 7.25,
|
| 2305 |
+
"rewards/rejected": -6.0,
|
| 2306 |
+
"step": 1500
|
| 2307 |
+
},
|
| 2308 |
+
{
|
| 2309 |
+
"epoch": 2.4198717948717947,
|
| 2310 |
+
"grad_norm": 13.945337786807688,
|
| 2311 |
+
"learning_rate": 1.0748218527315913e-07,
|
| 2312 |
+
"logits/chosen": -0.2353515625,
|
| 2313 |
+
"logits/rejected": -0.486328125,
|
| 2314 |
+
"logps/chosen": -374.0,
|
| 2315 |
+
"logps/rejected": -386.0,
|
| 2316 |
+
"loss": 0.0146,
|
| 2317 |
+
"rewards/accuracies": 1.0,
|
| 2318 |
+
"rewards/chosen": 2.28125,
|
| 2319 |
+
"rewards/margins": 8.625,
|
| 2320 |
+
"rewards/rejected": -6.34375,
|
| 2321 |
+
"step": 1510
|
| 2322 |
+
},
|
| 2323 |
+
{
|
| 2324 |
+
"epoch": 2.435897435897436,
|
| 2325 |
+
"grad_norm": 2.025562206086781,
|
| 2326 |
+
"learning_rate": 1.0451306413301662e-07,
|
| 2327 |
+
"logits/chosen": -0.2890625,
|
| 2328 |
+
"logits/rejected": -0.404296875,
|
| 2329 |
+
"logps/chosen": -362.0,
|
| 2330 |
+
"logps/rejected": -360.0,
|
| 2331 |
+
"loss": 0.0058,
|
| 2332 |
+
"rewards/accuracies": 1.0,
|
| 2333 |
+
"rewards/chosen": 2.96875,
|
| 2334 |
+
"rewards/margins": 8.875,
|
| 2335 |
+
"rewards/rejected": -5.90625,
|
| 2336 |
+
"step": 1520
|
| 2337 |
+
},
|
| 2338 |
+
{
|
| 2339 |
+
"epoch": 2.451923076923077,
|
| 2340 |
+
"grad_norm": 2.5818458781278997,
|
| 2341 |
+
"learning_rate": 1.0154394299287411e-07,
|
| 2342 |
+
"logits/chosen": -0.20703125,
|
| 2343 |
+
"logits/rejected": -0.380859375,
|
| 2344 |
+
"logps/chosen": -356.0,
|
| 2345 |
+
"logps/rejected": -378.0,
|
| 2346 |
+
"loss": 0.0084,
|
| 2347 |
+
"rewards/accuracies": 1.0,
|
| 2348 |
+
"rewards/chosen": 2.75,
|
| 2349 |
+
"rewards/margins": 8.375,
|
| 2350 |
+
"rewards/rejected": -5.59375,
|
| 2351 |
+
"step": 1530
|
| 2352 |
+
},
|
| 2353 |
+
{
|
| 2354 |
+
"epoch": 2.467948717948718,
|
| 2355 |
+
"grad_norm": 0.5147935469188503,
|
| 2356 |
+
"learning_rate": 9.857482185273158e-08,
|
| 2357 |
+
"logits/chosen": -0.240234375,
|
| 2358 |
+
"logits/rejected": -0.47265625,
|
| 2359 |
+
"logps/chosen": -348.0,
|
| 2360 |
+
"logps/rejected": -354.0,
|
| 2361 |
+
"loss": 0.0156,
|
| 2362 |
+
"rewards/accuracies": 1.0,
|
| 2363 |
+
"rewards/chosen": 2.609375,
|
| 2364 |
+
"rewards/margins": 8.125,
|
| 2365 |
+
"rewards/rejected": -5.46875,
|
| 2366 |
+
"step": 1540
|
| 2367 |
+
},
|
| 2368 |
+
{
|
| 2369 |
+
"epoch": 2.483974358974359,
|
| 2370 |
+
"grad_norm": 2.5527476910216764,
|
| 2371 |
+
"learning_rate": 9.560570071258907e-08,
|
| 2372 |
+
"logits/chosen": -0.1748046875,
|
| 2373 |
+
"logits/rejected": -0.392578125,
|
| 2374 |
+
"logps/chosen": -378.0,
|
| 2375 |
+
"logps/rejected": -336.0,
|
| 2376 |
+
"loss": 0.0132,
|
| 2377 |
+
"rewards/accuracies": 1.0,
|
| 2378 |
+
"rewards/chosen": 2.34375,
|
| 2379 |
+
"rewards/margins": 8.125,
|
| 2380 |
+
"rewards/rejected": -5.8125,
|
| 2381 |
+
"step": 1550
|
| 2382 |
+
},
|
| 2383 |
+
{
|
| 2384 |
+
"epoch": 2.5,
|
| 2385 |
+
"grad_norm": 2.6084669807533807,
|
| 2386 |
+
"learning_rate": 9.263657957244655e-08,
|
| 2387 |
+
"logits/chosen": -0.28125,
|
| 2388 |
+
"logits/rejected": -0.40234375,
|
| 2389 |
+
"logps/chosen": -340.0,
|
| 2390 |
+
"logps/rejected": -386.0,
|
| 2391 |
+
"loss": 0.0112,
|
| 2392 |
+
"rewards/accuracies": 0.987500011920929,
|
| 2393 |
+
"rewards/chosen": 2.78125,
|
| 2394 |
+
"rewards/margins": 8.4375,
|
| 2395 |
+
"rewards/rejected": -5.6875,
|
| 2396 |
+
"step": 1560
|
| 2397 |
+
},
|
| 2398 |
+
{
|
| 2399 |
+
"epoch": 2.516025641025641,
|
| 2400 |
+
"grad_norm": 0.5142214115866551,
|
| 2401 |
+
"learning_rate": 8.966745843230403e-08,
|
| 2402 |
+
"logits/chosen": -0.26953125,
|
| 2403 |
+
"logits/rejected": -0.34765625,
|
| 2404 |
+
"logps/chosen": -332.0,
|
| 2405 |
+
"logps/rejected": -338.0,
|
| 2406 |
+
"loss": 0.0087,
|
| 2407 |
+
"rewards/accuracies": 1.0,
|
| 2408 |
+
"rewards/chosen": 2.625,
|
| 2409 |
+
"rewards/margins": 8.25,
|
| 2410 |
+
"rewards/rejected": -5.625,
|
| 2411 |
+
"step": 1570
|
| 2412 |
+
},
|
| 2413 |
+
{
|
| 2414 |
+
"epoch": 2.532051282051282,
|
| 2415 |
+
"grad_norm": 0.4686133876015494,
|
| 2416 |
+
"learning_rate": 8.669833729216151e-08,
|
| 2417 |
+
"logits/chosen": -0.2080078125,
|
| 2418 |
+
"logits/rejected": -0.36328125,
|
| 2419 |
+
"logps/chosen": -374.0,
|
| 2420 |
+
"logps/rejected": -382.0,
|
| 2421 |
+
"loss": 0.0066,
|
| 2422 |
+
"rewards/accuracies": 1.0,
|
| 2423 |
+
"rewards/chosen": 2.0625,
|
| 2424 |
+
"rewards/margins": 7.84375,
|
| 2425 |
+
"rewards/rejected": -5.78125,
|
| 2426 |
+
"step": 1580
|
| 2427 |
+
},
|
| 2428 |
+
{
|
| 2429 |
+
"epoch": 2.5480769230769234,
|
| 2430 |
+
"grad_norm": 4.377458752115824,
|
| 2431 |
+
"learning_rate": 8.3729216152019e-08,
|
| 2432 |
+
"logits/chosen": -0.240234375,
|
| 2433 |
+
"logits/rejected": -0.50390625,
|
| 2434 |
+
"logps/chosen": -358.0,
|
| 2435 |
+
"logps/rejected": -370.0,
|
| 2436 |
+
"loss": 0.0077,
|
| 2437 |
+
"rewards/accuracies": 1.0,
|
| 2438 |
+
"rewards/chosen": 2.375,
|
| 2439 |
+
"rewards/margins": 8.4375,
|
| 2440 |
+
"rewards/rejected": -6.03125,
|
| 2441 |
+
"step": 1590
|
| 2442 |
+
},
|
| 2443 |
+
{
|
| 2444 |
+
"epoch": 2.564102564102564,
|
| 2445 |
+
"grad_norm": 8.575422939319253,
|
| 2446 |
+
"learning_rate": 8.076009501187649e-08,
|
| 2447 |
+
"logits/chosen": -0.2490234375,
|
| 2448 |
+
"logits/rejected": -0.337890625,
|
| 2449 |
+
"logps/chosen": -348.0,
|
| 2450 |
+
"logps/rejected": -370.0,
|
| 2451 |
+
"loss": 0.0241,
|
| 2452 |
+
"rewards/accuracies": 0.987500011920929,
|
| 2453 |
+
"rewards/chosen": 2.484375,
|
| 2454 |
+
"rewards/margins": 7.5,
|
| 2455 |
+
"rewards/rejected": -5.0,
|
| 2456 |
+
"step": 1600
|
| 2457 |
+
},
|
| 2458 |
+
{
|
| 2459 |
+
"epoch": 2.5801282051282053,
|
| 2460 |
+
"grad_norm": 2.551052943978534,
|
| 2461 |
+
"learning_rate": 7.779097387173396e-08,
|
| 2462 |
+
"logits/chosen": -0.2734375,
|
| 2463 |
+
"logits/rejected": -0.375,
|
| 2464 |
+
"logps/chosen": -342.0,
|
| 2465 |
+
"logps/rejected": -360.0,
|
| 2466 |
+
"loss": 0.0082,
|
| 2467 |
+
"rewards/accuracies": 1.0,
|
| 2468 |
+
"rewards/chosen": 2.8125,
|
| 2469 |
+
"rewards/margins": 7.8125,
|
| 2470 |
+
"rewards/rejected": -4.96875,
|
| 2471 |
+
"step": 1610
|
| 2472 |
+
},
|
| 2473 |
+
{
|
| 2474 |
+
"epoch": 2.5961538461538463,
|
| 2475 |
+
"grad_norm": 6.000172766868698,
|
| 2476 |
+
"learning_rate": 7.482185273159145e-08,
|
| 2477 |
+
"logits/chosen": -0.263671875,
|
| 2478 |
+
"logits/rejected": -0.451171875,
|
| 2479 |
+
"logps/chosen": -368.0,
|
| 2480 |
+
"logps/rejected": -378.0,
|
| 2481 |
+
"loss": 0.0303,
|
| 2482 |
+
"rewards/accuracies": 1.0,
|
| 2483 |
+
"rewards/chosen": 2.46875,
|
| 2484 |
+
"rewards/margins": 7.96875,
|
| 2485 |
+
"rewards/rejected": -5.5,
|
| 2486 |
+
"step": 1620
|
| 2487 |
+
},
|
| 2488 |
+
{
|
| 2489 |
+
"epoch": 2.6121794871794872,
|
| 2490 |
+
"grad_norm": 19.722033841376266,
|
| 2491 |
+
"learning_rate": 7.185273159144893e-08,
|
| 2492 |
+
"logits/chosen": -0.294921875,
|
| 2493 |
+
"logits/rejected": -0.392578125,
|
| 2494 |
+
"logps/chosen": -366.0,
|
| 2495 |
+
"logps/rejected": -398.0,
|
| 2496 |
+
"loss": 0.0123,
|
| 2497 |
+
"rewards/accuracies": 1.0,
|
| 2498 |
+
"rewards/chosen": 2.71875,
|
| 2499 |
+
"rewards/margins": 8.25,
|
| 2500 |
+
"rewards/rejected": -5.53125,
|
| 2501 |
+
"step": 1630
|
| 2502 |
+
},
|
| 2503 |
+
{
|
| 2504 |
+
"epoch": 2.628205128205128,
|
| 2505 |
+
"grad_norm": 11.30540085696599,
|
| 2506 |
+
"learning_rate": 6.88836104513064e-08,
|
| 2507 |
+
"logits/chosen": -0.18359375,
|
| 2508 |
+
"logits/rejected": -0.314453125,
|
| 2509 |
+
"logps/chosen": -380.0,
|
| 2510 |
+
"logps/rejected": -354.0,
|
| 2511 |
+
"loss": 0.0078,
|
| 2512 |
+
"rewards/accuracies": 1.0,
|
| 2513 |
+
"rewards/chosen": 2.734375,
|
| 2514 |
+
"rewards/margins": 8.6875,
|
| 2515 |
+
"rewards/rejected": -5.96875,
|
| 2516 |
+
"step": 1640
|
| 2517 |
+
},
|
| 2518 |
+
{
|
| 2519 |
+
"epoch": 2.644230769230769,
|
| 2520 |
+
"grad_norm": 1.1773482574998353,
|
| 2521 |
+
"learning_rate": 6.591448931116388e-08,
|
| 2522 |
+
"logits/chosen": -0.162109375,
|
| 2523 |
+
"logits/rejected": -0.4609375,
|
| 2524 |
+
"logps/chosen": -378.0,
|
| 2525 |
+
"logps/rejected": -374.0,
|
| 2526 |
+
"loss": 0.0175,
|
| 2527 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 2528 |
+
"rewards/chosen": 2.125,
|
| 2529 |
+
"rewards/margins": 8.0625,
|
| 2530 |
+
"rewards/rejected": -5.9375,
|
| 2531 |
+
"step": 1650
|
| 2532 |
+
},
|
| 2533 |
+
{
|
| 2534 |
+
"epoch": 2.66025641025641,
|
| 2535 |
+
"grad_norm": 52.67450880620409,
|
| 2536 |
+
"learning_rate": 6.294536817102138e-08,
|
| 2537 |
+
"logits/chosen": -0.294921875,
|
| 2538 |
+
"logits/rejected": -0.58203125,
|
| 2539 |
+
"logps/chosen": -386.0,
|
| 2540 |
+
"logps/rejected": -376.0,
|
| 2541 |
+
"loss": 0.0316,
|
| 2542 |
+
"rewards/accuracies": 0.987500011920929,
|
| 2543 |
+
"rewards/chosen": 1.828125,
|
| 2544 |
+
"rewards/margins": 7.53125,
|
| 2545 |
+
"rewards/rejected": -5.6875,
|
| 2546 |
+
"step": 1660
|
| 2547 |
+
},
|
| 2548 |
+
{
|
| 2549 |
+
"epoch": 2.676282051282051,
|
| 2550 |
+
"grad_norm": 2.444649208747581,
|
| 2551 |
+
"learning_rate": 5.997624703087885e-08,
|
| 2552 |
+
"logits/chosen": -0.2734375,
|
| 2553 |
+
"logits/rejected": -0.30859375,
|
| 2554 |
+
"logps/chosen": -348.0,
|
| 2555 |
+
"logps/rejected": -330.0,
|
| 2556 |
+
"loss": 0.0139,
|
| 2557 |
+
"rewards/accuracies": 1.0,
|
| 2558 |
+
"rewards/chosen": 2.34375,
|
| 2559 |
+
"rewards/margins": 7.96875,
|
| 2560 |
+
"rewards/rejected": -5.625,
|
| 2561 |
+
"step": 1670
|
| 2562 |
+
},
|
| 2563 |
+
{
|
| 2564 |
+
"epoch": 2.6923076923076925,
|
| 2565 |
+
"grad_norm": 1.8309710467577502,
|
| 2566 |
+
"learning_rate": 5.700712589073634e-08,
|
| 2567 |
+
"logits/chosen": -0.33203125,
|
| 2568 |
+
"logits/rejected": -0.5546875,
|
| 2569 |
+
"logps/chosen": -368.0,
|
| 2570 |
+
"logps/rejected": -372.0,
|
| 2571 |
+
"loss": 0.0044,
|
| 2572 |
+
"rewards/accuracies": 1.0,
|
| 2573 |
+
"rewards/chosen": 2.578125,
|
| 2574 |
+
"rewards/margins": 8.3125,
|
| 2575 |
+
"rewards/rejected": -5.75,
|
| 2576 |
+
"step": 1680
|
| 2577 |
+
},
|
| 2578 |
+
{
|
| 2579 |
+
"epoch": 2.7083333333333335,
|
| 2580 |
+
"grad_norm": 0.974183927512936,
|
| 2581 |
+
"learning_rate": 5.4038004750593824e-08,
|
| 2582 |
+
"logits/chosen": -0.212890625,
|
| 2583 |
+
"logits/rejected": -0.384765625,
|
| 2584 |
+
"logps/chosen": -370.0,
|
| 2585 |
+
"logps/rejected": -366.0,
|
| 2586 |
+
"loss": 0.0061,
|
| 2587 |
+
"rewards/accuracies": 1.0,
|
| 2588 |
+
"rewards/chosen": 2.4375,
|
| 2589 |
+
"rewards/margins": 8.625,
|
| 2590 |
+
"rewards/rejected": -6.21875,
|
| 2591 |
+
"step": 1690
|
| 2592 |
+
},
|
| 2593 |
+
{
|
| 2594 |
+
"epoch": 2.7243589743589745,
|
| 2595 |
+
"grad_norm": 8.112271810670284,
|
| 2596 |
+
"learning_rate": 5.10688836104513e-08,
|
| 2597 |
+
"logits/chosen": -0.287109375,
|
| 2598 |
+
"logits/rejected": -0.34375,
|
| 2599 |
+
"logps/chosen": -364.0,
|
| 2600 |
+
"logps/rejected": -384.0,
|
| 2601 |
+
"loss": 0.0074,
|
| 2602 |
+
"rewards/accuracies": 1.0,
|
| 2603 |
+
"rewards/chosen": 2.53125,
|
| 2604 |
+
"rewards/margins": 8.625,
|
| 2605 |
+
"rewards/rejected": -6.09375,
|
| 2606 |
+
"step": 1700
|
| 2607 |
+
},
|
| 2608 |
+
{
|
| 2609 |
+
"epoch": 2.7403846153846154,
|
| 2610 |
+
"grad_norm": 48.323303487774965,
|
| 2611 |
+
"learning_rate": 4.809976247030879e-08,
|
| 2612 |
+
"logits/chosen": -0.2734375,
|
| 2613 |
+
"logits/rejected": -0.39453125,
|
| 2614 |
+
"logps/chosen": -378.0,
|
| 2615 |
+
"logps/rejected": -368.0,
|
| 2616 |
+
"loss": 0.0187,
|
| 2617 |
+
"rewards/accuracies": 1.0,
|
| 2618 |
+
"rewards/chosen": 2.28125,
|
| 2619 |
+
"rewards/margins": 7.9375,
|
| 2620 |
+
"rewards/rejected": -5.65625,
|
| 2621 |
+
"step": 1710
|
| 2622 |
+
},
|
| 2623 |
+
{
|
| 2624 |
+
"epoch": 2.7564102564102564,
|
| 2625 |
+
"grad_norm": 0.6216279700601488,
|
| 2626 |
+
"learning_rate": 4.5130641330166267e-08,
|
| 2627 |
+
"logits/chosen": -0.2470703125,
|
| 2628 |
+
"logits/rejected": -0.3828125,
|
| 2629 |
+
"logps/chosen": -340.0,
|
| 2630 |
+
"logps/rejected": -372.0,
|
| 2631 |
+
"loss": 0.0305,
|
| 2632 |
+
"rewards/accuracies": 0.987500011920929,
|
| 2633 |
+
"rewards/chosen": 2.25,
|
| 2634 |
+
"rewards/margins": 7.6875,
|
| 2635 |
+
"rewards/rejected": -5.4375,
|
| 2636 |
+
"step": 1720
|
| 2637 |
+
},
|
| 2638 |
+
{
|
| 2639 |
+
"epoch": 2.7724358974358974,
|
| 2640 |
+
"grad_norm": 8.386988307728698,
|
| 2641 |
+
"learning_rate": 4.216152019002375e-08,
|
| 2642 |
+
"logits/chosen": -0.2060546875,
|
| 2643 |
+
"logits/rejected": -0.349609375,
|
| 2644 |
+
"logps/chosen": -360.0,
|
| 2645 |
+
"logps/rejected": -364.0,
|
| 2646 |
+
"loss": 0.0062,
|
| 2647 |
+
"rewards/accuracies": 1.0,
|
| 2648 |
+
"rewards/chosen": 2.734375,
|
| 2649 |
+
"rewards/margins": 8.4375,
|
| 2650 |
+
"rewards/rejected": -5.6875,
|
| 2651 |
+
"step": 1730
|
| 2652 |
+
},
|
| 2653 |
+
{
|
| 2654 |
+
"epoch": 2.7884615384615383,
|
| 2655 |
+
"grad_norm": 2.6310329285858587,
|
| 2656 |
+
"learning_rate": 3.919239904988123e-08,
|
| 2657 |
+
"logits/chosen": -0.1826171875,
|
| 2658 |
+
"logits/rejected": -0.333984375,
|
| 2659 |
+
"logps/chosen": -358.0,
|
| 2660 |
+
"logps/rejected": -390.0,
|
| 2661 |
+
"loss": 0.0088,
|
| 2662 |
+
"rewards/accuracies": 1.0,
|
| 2663 |
+
"rewards/chosen": 2.28125,
|
| 2664 |
+
"rewards/margins": 7.875,
|
| 2665 |
+
"rewards/rejected": -5.59375,
|
| 2666 |
+
"step": 1740
|
| 2667 |
+
},
|
| 2668 |
+
{
|
| 2669 |
+
"epoch": 2.8044871794871797,
|
| 2670 |
+
"grad_norm": 3.066349322026837,
|
| 2671 |
+
"learning_rate": 3.6223277909738715e-08,
|
| 2672 |
+
"logits/chosen": -0.2236328125,
|
| 2673 |
+
"logits/rejected": -0.357421875,
|
| 2674 |
+
"logps/chosen": -350.0,
|
| 2675 |
+
"logps/rejected": -368.0,
|
| 2676 |
+
"loss": 0.0092,
|
| 2677 |
+
"rewards/accuracies": 1.0,
|
| 2678 |
+
"rewards/chosen": 2.40625,
|
| 2679 |
+
"rewards/margins": 7.96875,
|
| 2680 |
+
"rewards/rejected": -5.5625,
|
| 2681 |
+
"step": 1750
|
| 2682 |
+
},
|
| 2683 |
+
{
|
| 2684 |
+
"epoch": 2.8205128205128203,
|
| 2685 |
+
"grad_norm": 8.079242737705172,
|
| 2686 |
+
"learning_rate": 3.32541567695962e-08,
|
| 2687 |
+
"logits/chosen": -0.0966796875,
|
| 2688 |
+
"logits/rejected": -0.2431640625,
|
| 2689 |
+
"logps/chosen": -342.0,
|
| 2690 |
+
"logps/rejected": -360.0,
|
| 2691 |
+
"loss": 0.0237,
|
| 2692 |
+
"rewards/accuracies": 1.0,
|
| 2693 |
+
"rewards/chosen": 2.421875,
|
| 2694 |
+
"rewards/margins": 8.4375,
|
| 2695 |
+
"rewards/rejected": -6.03125,
|
| 2696 |
+
"step": 1760
|
| 2697 |
+
},
|
| 2698 |
+
{
|
| 2699 |
+
"epoch": 2.8365384615384617,
|
| 2700 |
+
"grad_norm": 36.85905904660706,
|
| 2701 |
+
"learning_rate": 3.028503562945368e-08,
|
| 2702 |
+
"logits/chosen": -0.1962890625,
|
| 2703 |
+
"logits/rejected": -0.1435546875,
|
| 2704 |
+
"logps/chosen": -376.0,
|
| 2705 |
+
"logps/rejected": -366.0,
|
| 2706 |
+
"loss": 0.0138,
|
| 2707 |
+
"rewards/accuracies": 1.0,
|
| 2708 |
+
"rewards/chosen": 3.015625,
|
| 2709 |
+
"rewards/margins": 8.75,
|
| 2710 |
+
"rewards/rejected": -5.71875,
|
| 2711 |
+
"step": 1770
|
| 2712 |
+
},
|
| 2713 |
+
{
|
| 2714 |
+
"epoch": 2.8525641025641026,
|
| 2715 |
+
"grad_norm": 1.3700769518291278,
|
| 2716 |
+
"learning_rate": 2.7315914489311164e-08,
|
| 2717 |
+
"logits/chosen": -0.177734375,
|
| 2718 |
+
"logits/rejected": -0.29296875,
|
| 2719 |
+
"logps/chosen": -354.0,
|
| 2720 |
+
"logps/rejected": -366.0,
|
| 2721 |
+
"loss": 0.0047,
|
| 2722 |
+
"rewards/accuracies": 1.0,
|
| 2723 |
+
"rewards/chosen": 2.1875,
|
| 2724 |
+
"rewards/margins": 8.375,
|
| 2725 |
+
"rewards/rejected": -6.15625,
|
| 2726 |
+
"step": 1780
|
| 2727 |
+
},
|
| 2728 |
+
{
|
| 2729 |
+
"epoch": 2.8685897435897436,
|
| 2730 |
+
"grad_norm": 2.5772413830882437,
|
| 2731 |
+
"learning_rate": 2.4346793349168646e-08,
|
| 2732 |
+
"logits/chosen": -0.3515625,
|
| 2733 |
+
"logits/rejected": -0.40625,
|
| 2734 |
+
"logps/chosen": -356.0,
|
| 2735 |
+
"logps/rejected": -380.0,
|
| 2736 |
+
"loss": 0.0071,
|
| 2737 |
+
"rewards/accuracies": 1.0,
|
| 2738 |
+
"rewards/chosen": 2.265625,
|
| 2739 |
+
"rewards/margins": 8.25,
|
| 2740 |
+
"rewards/rejected": -5.96875,
|
| 2741 |
+
"step": 1790
|
| 2742 |
+
},
|
| 2743 |
+
{
|
| 2744 |
+
"epoch": 2.8846153846153846,
|
| 2745 |
+
"grad_norm": 0.2522421688634616,
|
| 2746 |
+
"learning_rate": 2.1377672209026125e-08,
|
| 2747 |
+
"logits/chosen": -0.1884765625,
|
| 2748 |
+
"logits/rejected": -0.296875,
|
| 2749 |
+
"logps/chosen": -354.0,
|
| 2750 |
+
"logps/rejected": -362.0,
|
| 2751 |
+
"loss": 0.0155,
|
| 2752 |
+
"rewards/accuracies": 0.987500011920929,
|
| 2753 |
+
"rewards/chosen": 2.609375,
|
| 2754 |
+
"rewards/margins": 8.625,
|
| 2755 |
+
"rewards/rejected": -5.96875,
|
| 2756 |
+
"step": 1800
|
| 2757 |
+
},
|
| 2758 |
+
{
|
| 2759 |
+
"epoch": 2.9006410256410255,
|
| 2760 |
+
"grad_norm": 2.0378946111088347,
|
| 2761 |
+
"learning_rate": 1.840855106888361e-08,
|
| 2762 |
+
"logits/chosen": -0.298828125,
|
| 2763 |
+
"logits/rejected": -0.4765625,
|
| 2764 |
+
"logps/chosen": -346.0,
|
| 2765 |
+
"logps/rejected": -370.0,
|
| 2766 |
+
"loss": 0.0427,
|
| 2767 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 2768 |
+
"rewards/chosen": 1.828125,
|
| 2769 |
+
"rewards/margins": 7.90625,
|
| 2770 |
+
"rewards/rejected": -6.0625,
|
| 2771 |
+
"step": 1810
|
| 2772 |
+
},
|
| 2773 |
+
{
|
| 2774 |
+
"epoch": 2.9166666666666665,
|
| 2775 |
+
"grad_norm": 2.2386092025677784,
|
| 2776 |
+
"learning_rate": 1.5439429928741092e-08,
|
| 2777 |
+
"logits/chosen": -0.21875,
|
| 2778 |
+
"logits/rejected": -0.2138671875,
|
| 2779 |
+
"logps/chosen": -380.0,
|
| 2780 |
+
"logps/rejected": -372.0,
|
| 2781 |
+
"loss": 0.0179,
|
| 2782 |
+
"rewards/accuracies": 0.987500011920929,
|
| 2783 |
+
"rewards/chosen": 1.84375,
|
| 2784 |
+
"rewards/margins": 8.5,
|
| 2785 |
+
"rewards/rejected": -6.65625,
|
| 2786 |
+
"step": 1820
|
| 2787 |
+
},
|
| 2788 |
+
{
|
| 2789 |
+
"epoch": 2.9326923076923075,
|
| 2790 |
+
"grad_norm": 5.606669828245322,
|
| 2791 |
+
"learning_rate": 1.2470308788598574e-08,
|
| 2792 |
+
"logits/chosen": -0.26953125,
|
| 2793 |
+
"logits/rejected": -0.224609375,
|
| 2794 |
+
"logps/chosen": -364.0,
|
| 2795 |
+
"logps/rejected": -346.0,
|
| 2796 |
+
"loss": 0.0302,
|
| 2797 |
+
"rewards/accuracies": 1.0,
|
| 2798 |
+
"rewards/chosen": 2.5625,
|
| 2799 |
+
"rewards/margins": 8.5625,
|
| 2800 |
+
"rewards/rejected": -5.96875,
|
| 2801 |
+
"step": 1830
|
| 2802 |
+
},
|
| 2803 |
+
{
|
| 2804 |
+
"epoch": 2.948717948717949,
|
| 2805 |
+
"grad_norm": 7.459897331240726,
|
| 2806 |
+
"learning_rate": 9.501187648456057e-09,
|
| 2807 |
+
"logits/chosen": -0.166015625,
|
| 2808 |
+
"logits/rejected": -0.27734375,
|
| 2809 |
+
"logps/chosen": -338.0,
|
| 2810 |
+
"logps/rejected": -372.0,
|
| 2811 |
+
"loss": 0.0069,
|
| 2812 |
+
"rewards/accuracies": 1.0,
|
| 2813 |
+
"rewards/chosen": 1.9453125,
|
| 2814 |
+
"rewards/margins": 8.375,
|
| 2815 |
+
"rewards/rejected": -6.40625,
|
| 2816 |
+
"step": 1840
|
| 2817 |
+
},
|
| 2818 |
+
{
|
| 2819 |
+
"epoch": 2.96474358974359,
|
| 2820 |
+
"grad_norm": 46.309709716339626,
|
| 2821 |
+
"learning_rate": 6.532066508313539e-09,
|
| 2822 |
+
"logits/chosen": -0.265625,
|
| 2823 |
+
"logits/rejected": -0.3046875,
|
| 2824 |
+
"logps/chosen": -368.0,
|
| 2825 |
+
"logps/rejected": -392.0,
|
| 2826 |
+
"loss": 0.0144,
|
| 2827 |
+
"rewards/accuracies": 1.0,
|
| 2828 |
+
"rewards/chosen": 2.125,
|
| 2829 |
+
"rewards/margins": 8.5,
|
| 2830 |
+
"rewards/rejected": -6.375,
|
| 2831 |
+
"step": 1850
|
| 2832 |
+
},
|
| 2833 |
+
{
|
| 2834 |
+
"epoch": 2.980769230769231,
|
| 2835 |
+
"grad_norm": 2.757504775523463,
|
| 2836 |
+
"learning_rate": 3.562945368171021e-09,
|
| 2837 |
+
"logits/chosen": -0.2314453125,
|
| 2838 |
+
"logits/rejected": -0.3515625,
|
| 2839 |
+
"logps/chosen": -366.0,
|
| 2840 |
+
"logps/rejected": -382.0,
|
| 2841 |
+
"loss": 0.0081,
|
| 2842 |
+
"rewards/accuracies": 1.0,
|
| 2843 |
+
"rewards/chosen": 2.3125,
|
| 2844 |
+
"rewards/margins": 8.125,
|
| 2845 |
+
"rewards/rejected": -5.84375,
|
| 2846 |
+
"step": 1860
|
| 2847 |
+
},
|
| 2848 |
+
{
|
| 2849 |
+
"epoch": 2.996794871794872,
|
| 2850 |
+
"grad_norm": 0.7690916064350508,
|
| 2851 |
+
"learning_rate": 5.938242280285036e-10,
|
| 2852 |
+
"logits/chosen": -0.146484375,
|
| 2853 |
+
"logits/rejected": -0.361328125,
|
| 2854 |
+
"logps/chosen": -352.0,
|
| 2855 |
+
"logps/rejected": -378.0,
|
| 2856 |
+
"loss": 0.0214,
|
| 2857 |
+
"rewards/accuracies": 1.0,
|
| 2858 |
+
"rewards/chosen": 2.9375,
|
| 2859 |
+
"rewards/margins": 8.75,
|
| 2860 |
+
"rewards/rejected": -5.84375,
|
| 2861 |
+
"step": 1870
|
| 2862 |
+
},
|
| 2863 |
+
{
|
| 2864 |
+
"epoch": 3.0,
|
| 2865 |
+
"eval_logits/chosen": -0.2392578125,
|
| 2866 |
+
"eval_logits/rejected": -0.2734375,
|
| 2867 |
+
"eval_logps/chosen": -364.0,
|
| 2868 |
+
"eval_logps/rejected": -370.0,
|
| 2869 |
+
"eval_loss": 0.21664032340049744,
|
| 2870 |
+
"eval_rewards/accuracies": 0.9134615659713745,
|
| 2871 |
+
"eval_rewards/chosen": 1.7734375,
|
| 2872 |
+
"eval_rewards/margins": 6.65625,
|
| 2873 |
+
"eval_rewards/rejected": -4.875,
|
| 2874 |
+
"eval_runtime": 26.3085,
|
| 2875 |
+
"eval_samples_per_second": 7.602,
|
| 2876 |
+
"eval_steps_per_second": 0.494,
|
| 2877 |
+
"step": 1872
|
| 2878 |
+
},
|
| 2879 |
+
{
|
| 2880 |
+
"epoch": 3.0,
|
| 2881 |
+
"step": 1872,
|
| 2882 |
+
"total_flos": 0.0,
|
| 2883 |
+
"train_loss": 0.14497800362415802,
|
| 2884 |
+
"train_runtime": 10901.8033,
|
| 2885 |
+
"train_samples_per_second": 2.746,
|
| 2886 |
+
"train_steps_per_second": 0.172
|
| 2887 |
+
}
|
| 2888 |
+
],
|
| 2889 |
+
"logging_steps": 10,
|
| 2890 |
+
"max_steps": 1872,
|
| 2891 |
+
"num_input_tokens_seen": 0,
|
| 2892 |
+
"num_train_epochs": 3,
|
| 2893 |
+
"save_steps": 500,
|
| 2894 |
+
"stateful_callbacks": {
|
| 2895 |
+
"TrainerControl": {
|
| 2896 |
+
"args": {
|
| 2897 |
+
"should_epoch_stop": false,
|
| 2898 |
+
"should_evaluate": false,
|
| 2899 |
+
"should_log": false,
|
| 2900 |
+
"should_save": true,
|
| 2901 |
+
"should_training_stop": true
|
| 2902 |
+
},
|
| 2903 |
+
"attributes": {}
|
| 2904 |
+
}
|
| 2905 |
+
},
|
| 2906 |
+
"total_flos": 0.0,
|
| 2907 |
+
"train_batch_size": 8,
|
| 2908 |
+
"trial_name": null,
|
| 2909 |
+
"trial_params": null
|
| 2910 |
+
}
|