Model save
Browse files
README.md
ADDED
|
@@ -0,0 +1,68 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
base_model: philschmid/llama-3-1-8b-math-orca-spectrum-10k-ep1
|
| 3 |
+
library_name: transformers
|
| 4 |
+
model_name: grpo-llama-3-1-8b-math-ep3
|
| 5 |
+
tags:
|
| 6 |
+
- generated_from_trainer
|
| 7 |
+
- trl
|
| 8 |
+
- grpo
|
| 9 |
+
licence: license
|
| 10 |
+
---
|
| 11 |
+
|
| 12 |
+
# Model Card for grpo-llama-3-1-8b-math-ep3
|
| 13 |
+
|
| 14 |
+
This model is a fine-tuned version of [philschmid/llama-3-1-8b-math-orca-spectrum-10k-ep1](https://huggingface.co/philschmid/llama-3-1-8b-math-orca-spectrum-10k-ep1).
|
| 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="nghind/grpo-llama-3-1-8b-math-ep3", 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 |
+
|
| 31 |
+
|
| 32 |
+
|
| 33 |
+
This model was trained with GRPO, a method introduced in [DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models](https://huggingface.co/papers/2402.03300).
|
| 34 |
+
|
| 35 |
+
### Framework versions
|
| 36 |
+
|
| 37 |
+
- TRL: 0.15.1
|
| 38 |
+
- Transformers: 4.49.0
|
| 39 |
+
- Pytorch: 2.5.1+cu121
|
| 40 |
+
- Datasets: 3.3.1
|
| 41 |
+
- Tokenizers: 0.21.0
|
| 42 |
+
|
| 43 |
+
## Citations
|
| 44 |
+
|
| 45 |
+
Cite GRPO as:
|
| 46 |
+
|
| 47 |
+
```bibtex
|
| 48 |
+
@article{zhihong2024deepseekmath,
|
| 49 |
+
title = {{DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models}},
|
| 50 |
+
author = {Zhihong Shao and Peiyi Wang and Qihao Zhu and Runxin Xu and Junxiao Song and Mingchuan Zhang and Y. K. Li and Y. Wu and Daya Guo},
|
| 51 |
+
year = 2024,
|
| 52 |
+
eprint = {arXiv:2402.03300},
|
| 53 |
+
}
|
| 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 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"total_flos": 0.0,
|
| 3 |
+
"train_loss": 0.00012346780326337724,
|
| 4 |
+
"train_runtime": 38700.7369,
|
| 5 |
+
"train_samples": 7943,
|
| 6 |
+
"train_samples_per_second": 0.616,
|
| 7 |
+
"train_steps_per_second": 0.019
|
| 8 |
+
}
|
runs/Feb20_10-08-19_koa-dgxa-b11-u17/events.out.tfevents.1740046206.koa-dgxa-b11-u17.2389931.0
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:98cb593a614ca4c271cac54520d34834d166f4f4a1dd2e73a610c9da1b81c0de
|
| 3 |
+
size 80356
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"total_flos": 0.0,
|
| 3 |
+
"train_loss": 0.00012346780326337724,
|
| 4 |
+
"train_runtime": 38700.7369,
|
| 5 |
+
"train_samples": 7943,
|
| 6 |
+
"train_samples_per_second": 0.616,
|
| 7 |
+
"train_steps_per_second": 0.019
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1835 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 3.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 747,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"completion_length": 159.95703125,
|
| 13 |
+
"epoch": 0.020080321285140562,
|
| 14 |
+
"grad_norm": 0.2833329439163208,
|
| 15 |
+
"kl": 0.00034067649394273756,
|
| 16 |
+
"learning_rate": 5e-06,
|
| 17 |
+
"loss": 0.0,
|
| 18 |
+
"reward": 0.32109375,
|
| 19 |
+
"reward_std": 0.2864044725894928,
|
| 20 |
+
"rewards/acc_reward_func": 0.32109375,
|
| 21 |
+
"step": 5
|
| 22 |
+
},
|
| 23 |
+
{
|
| 24 |
+
"completion_length": 179.0203125,
|
| 25 |
+
"epoch": 0.040160642570281124,
|
| 26 |
+
"grad_norm": 0.47351399064064026,
|
| 27 |
+
"kl": 0.000714331166818738,
|
| 28 |
+
"learning_rate": 5e-06,
|
| 29 |
+
"loss": 0.0001,
|
| 30 |
+
"reward": 0.30625,
|
| 31 |
+
"reward_std": 0.2533454954624176,
|
| 32 |
+
"rewards/acc_reward_func": 0.30625,
|
| 33 |
+
"step": 10
|
| 34 |
+
},
|
| 35 |
+
{
|
| 36 |
+
"completion_length": 163.7078125,
|
| 37 |
+
"epoch": 0.060240963855421686,
|
| 38 |
+
"grad_norm": 0.6223832368850708,
|
| 39 |
+
"kl": 0.0008577127242460847,
|
| 40 |
+
"learning_rate": 5e-06,
|
| 41 |
+
"loss": 0.0001,
|
| 42 |
+
"reward": 0.290625,
|
| 43 |
+
"reward_std": 0.26476518511772157,
|
| 44 |
+
"rewards/acc_reward_func": 0.290625,
|
| 45 |
+
"step": 15
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"completion_length": 177.18203125,
|
| 49 |
+
"epoch": 0.08032128514056225,
|
| 50 |
+
"grad_norm": 0.3723163902759552,
|
| 51 |
+
"kl": 0.000897675973828882,
|
| 52 |
+
"learning_rate": 5e-06,
|
| 53 |
+
"loss": 0.0001,
|
| 54 |
+
"reward": 0.33125,
|
| 55 |
+
"reward_std": 0.27346172034740446,
|
| 56 |
+
"rewards/acc_reward_func": 0.33125,
|
| 57 |
+
"step": 20
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"completion_length": 172.44921875,
|
| 61 |
+
"epoch": 0.10040160642570281,
|
| 62 |
+
"grad_norm": 0.4859201908111572,
|
| 63 |
+
"kl": 0.0007807362941093743,
|
| 64 |
+
"learning_rate": 5e-06,
|
| 65 |
+
"loss": 0.0001,
|
| 66 |
+
"reward": 0.35625,
|
| 67 |
+
"reward_std": 0.28146005868911744,
|
| 68 |
+
"rewards/acc_reward_func": 0.35625,
|
| 69 |
+
"step": 25
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"completion_length": 173.3265625,
|
| 73 |
+
"epoch": 0.12048192771084337,
|
| 74 |
+
"grad_norm": 0.4038971960544586,
|
| 75 |
+
"kl": 0.0007229511742480099,
|
| 76 |
+
"learning_rate": 5e-06,
|
| 77 |
+
"loss": 0.0001,
|
| 78 |
+
"reward": 0.31640625,
|
| 79 |
+
"reward_std": 0.2735124319791794,
|
| 80 |
+
"rewards/acc_reward_func": 0.31640625,
|
| 81 |
+
"step": 30
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"completion_length": 168.64296875,
|
| 85 |
+
"epoch": 0.14056224899598393,
|
| 86 |
+
"grad_norm": 0.3058616816997528,
|
| 87 |
+
"kl": 0.000607735151425004,
|
| 88 |
+
"learning_rate": 5e-06,
|
| 89 |
+
"loss": 0.0001,
|
| 90 |
+
"reward": 0.3265625,
|
| 91 |
+
"reward_std": 0.24166617095470427,
|
| 92 |
+
"rewards/acc_reward_func": 0.3265625,
|
| 93 |
+
"step": 35
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"completion_length": 162.60703125,
|
| 97 |
+
"epoch": 0.1606425702811245,
|
| 98 |
+
"grad_norm": 0.2520759701728821,
|
| 99 |
+
"kl": 0.0005803823471069336,
|
| 100 |
+
"learning_rate": 5e-06,
|
| 101 |
+
"loss": 0.0001,
|
| 102 |
+
"reward": 0.309375,
|
| 103 |
+
"reward_std": 0.2683109283447266,
|
| 104 |
+
"rewards/acc_reward_func": 0.309375,
|
| 105 |
+
"step": 40
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"completion_length": 167.0765625,
|
| 109 |
+
"epoch": 0.18072289156626506,
|
| 110 |
+
"grad_norm": 0.3606955409049988,
|
| 111 |
+
"kl": 0.00041882623336277904,
|
| 112 |
+
"learning_rate": 5e-06,
|
| 113 |
+
"loss": 0.0,
|
| 114 |
+
"reward": 0.3390625,
|
| 115 |
+
"reward_std": 0.270548814535141,
|
| 116 |
+
"rewards/acc_reward_func": 0.3390625,
|
| 117 |
+
"step": 45
|
| 118 |
+
},
|
| 119 |
+
{
|
| 120 |
+
"completion_length": 173.20859375,
|
| 121 |
+
"epoch": 0.20080321285140562,
|
| 122 |
+
"grad_norm": 0.43256309628486633,
|
| 123 |
+
"kl": 0.0004548234341200441,
|
| 124 |
+
"learning_rate": 5e-06,
|
| 125 |
+
"loss": 0.0,
|
| 126 |
+
"reward": 0.3125,
|
| 127 |
+
"reward_std": 0.24177098274230957,
|
| 128 |
+
"rewards/acc_reward_func": 0.3125,
|
| 129 |
+
"step": 50
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"completion_length": 168.24296875,
|
| 133 |
+
"epoch": 0.22088353413654618,
|
| 134 |
+
"grad_norm": 0.5047885775566101,
|
| 135 |
+
"kl": 0.0004514524363912642,
|
| 136 |
+
"learning_rate": 5e-06,
|
| 137 |
+
"loss": 0.0,
|
| 138 |
+
"reward": 0.35,
|
| 139 |
+
"reward_std": 0.27911658585071564,
|
| 140 |
+
"rewards/acc_reward_func": 0.35,
|
| 141 |
+
"step": 55
|
| 142 |
+
},
|
| 143 |
+
{
|
| 144 |
+
"completion_length": 168.32109375,
|
| 145 |
+
"epoch": 0.24096385542168675,
|
| 146 |
+
"grad_norm": 0.2552633285522461,
|
| 147 |
+
"kl": 0.00045427558943629266,
|
| 148 |
+
"learning_rate": 5e-06,
|
| 149 |
+
"loss": 0.0,
|
| 150 |
+
"reward": 0.31015625,
|
| 151 |
+
"reward_std": 0.27751815915107725,
|
| 152 |
+
"rewards/acc_reward_func": 0.31015625,
|
| 153 |
+
"step": 60
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"completion_length": 173.72421875,
|
| 157 |
+
"epoch": 0.26104417670682734,
|
| 158 |
+
"grad_norm": 0.2749291956424713,
|
| 159 |
+
"kl": 0.0004708627995569259,
|
| 160 |
+
"learning_rate": 5e-06,
|
| 161 |
+
"loss": 0.0,
|
| 162 |
+
"reward": 0.36171875,
|
| 163 |
+
"reward_std": 0.2908409178256989,
|
| 164 |
+
"rewards/acc_reward_func": 0.36171875,
|
| 165 |
+
"step": 65
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"completion_length": 174.65390625,
|
| 169 |
+
"epoch": 0.28112449799196787,
|
| 170 |
+
"grad_norm": 0.26198047399520874,
|
| 171 |
+
"kl": 0.000500894442666322,
|
| 172 |
+
"learning_rate": 5e-06,
|
| 173 |
+
"loss": 0.0001,
|
| 174 |
+
"reward": 0.29296875,
|
| 175 |
+
"reward_std": 0.2530863583087921,
|
| 176 |
+
"rewards/acc_reward_func": 0.29296875,
|
| 177 |
+
"step": 70
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"completion_length": 168.96953125,
|
| 181 |
+
"epoch": 0.30120481927710846,
|
| 182 |
+
"grad_norm": 0.34051886200904846,
|
| 183 |
+
"kl": 0.0004913369542919099,
|
| 184 |
+
"learning_rate": 5e-06,
|
| 185 |
+
"loss": 0.0,
|
| 186 |
+
"reward": 0.34453125,
|
| 187 |
+
"reward_std": 0.2887667536735535,
|
| 188 |
+
"rewards/acc_reward_func": 0.34453125,
|
| 189 |
+
"step": 75
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"completion_length": 166.57109375,
|
| 193 |
+
"epoch": 0.321285140562249,
|
| 194 |
+
"grad_norm": 0.2959311306476593,
|
| 195 |
+
"kl": 0.0004982782644219697,
|
| 196 |
+
"learning_rate": 5e-06,
|
| 197 |
+
"loss": 0.0,
|
| 198 |
+
"reward": 0.328125,
|
| 199 |
+
"reward_std": 0.25415654480457306,
|
| 200 |
+
"rewards/acc_reward_func": 0.328125,
|
| 201 |
+
"step": 80
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"completion_length": 169.40703125,
|
| 205 |
+
"epoch": 0.3413654618473896,
|
| 206 |
+
"grad_norm": 0.26805025339126587,
|
| 207 |
+
"kl": 0.0005323876044712961,
|
| 208 |
+
"learning_rate": 5e-06,
|
| 209 |
+
"loss": 0.0001,
|
| 210 |
+
"reward": 0.36796875,
|
| 211 |
+
"reward_std": 0.2889047384262085,
|
| 212 |
+
"rewards/acc_reward_func": 0.36796875,
|
| 213 |
+
"step": 85
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"completion_length": 167.22890625,
|
| 217 |
+
"epoch": 0.3614457831325301,
|
| 218 |
+
"grad_norm": 0.3810369074344635,
|
| 219 |
+
"kl": 0.0004968916124198586,
|
| 220 |
+
"learning_rate": 5e-06,
|
| 221 |
+
"loss": 0.0,
|
| 222 |
+
"reward": 0.36015625,
|
| 223 |
+
"reward_std": 0.2771797090768814,
|
| 224 |
+
"rewards/acc_reward_func": 0.36015625,
|
| 225 |
+
"step": 90
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"completion_length": 175.2796875,
|
| 229 |
+
"epoch": 0.3815261044176707,
|
| 230 |
+
"grad_norm": 0.28440332412719727,
|
| 231 |
+
"kl": 0.0004675893171224743,
|
| 232 |
+
"learning_rate": 5e-06,
|
| 233 |
+
"loss": 0.0,
|
| 234 |
+
"reward": 0.31484375,
|
| 235 |
+
"reward_std": 0.27064948081970214,
|
| 236 |
+
"rewards/acc_reward_func": 0.31484375,
|
| 237 |
+
"step": 95
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"completion_length": 163.48203125,
|
| 241 |
+
"epoch": 0.40160642570281124,
|
| 242 |
+
"grad_norm": 0.3027651309967041,
|
| 243 |
+
"kl": 0.0004759302770253271,
|
| 244 |
+
"learning_rate": 5e-06,
|
| 245 |
+
"loss": 0.0,
|
| 246 |
+
"reward": 0.3125,
|
| 247 |
+
"reward_std": 0.2688873440027237,
|
| 248 |
+
"rewards/acc_reward_func": 0.3125,
|
| 249 |
+
"step": 100
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"completion_length": 175.22734375,
|
| 253 |
+
"epoch": 0.42168674698795183,
|
| 254 |
+
"grad_norm": 0.5133289098739624,
|
| 255 |
+
"kl": 0.0004573037032969296,
|
| 256 |
+
"learning_rate": 5e-06,
|
| 257 |
+
"loss": 0.0,
|
| 258 |
+
"reward": 0.31015625,
|
| 259 |
+
"reward_std": 0.2805974006652832,
|
| 260 |
+
"rewards/acc_reward_func": 0.31015625,
|
| 261 |
+
"step": 105
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"completion_length": 167.0984375,
|
| 265 |
+
"epoch": 0.44176706827309237,
|
| 266 |
+
"grad_norm": 0.3459770381450653,
|
| 267 |
+
"kl": 0.0005908279563300312,
|
| 268 |
+
"learning_rate": 5e-06,
|
| 269 |
+
"loss": 0.0001,
|
| 270 |
+
"reward": 0.33359375,
|
| 271 |
+
"reward_std": 0.2592057645320892,
|
| 272 |
+
"rewards/acc_reward_func": 0.33359375,
|
| 273 |
+
"step": 110
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"completion_length": 167.68671875,
|
| 277 |
+
"epoch": 0.46184738955823296,
|
| 278 |
+
"grad_norm": 0.4075464606285095,
|
| 279 |
+
"kl": 0.0006865185336209833,
|
| 280 |
+
"learning_rate": 5e-06,
|
| 281 |
+
"loss": 0.0001,
|
| 282 |
+
"reward": 0.38203125,
|
| 283 |
+
"reward_std": 0.265577495098114,
|
| 284 |
+
"rewards/acc_reward_func": 0.38203125,
|
| 285 |
+
"step": 115
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"completion_length": 163.8390625,
|
| 289 |
+
"epoch": 0.4819277108433735,
|
| 290 |
+
"grad_norm": 0.37416180968284607,
|
| 291 |
+
"kl": 0.000659433496184647,
|
| 292 |
+
"learning_rate": 5e-06,
|
| 293 |
+
"loss": 0.0001,
|
| 294 |
+
"reward": 0.315625,
|
| 295 |
+
"reward_std": 0.2419831484556198,
|
| 296 |
+
"rewards/acc_reward_func": 0.315625,
|
| 297 |
+
"step": 120
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"completion_length": 167.946875,
|
| 301 |
+
"epoch": 0.5020080321285141,
|
| 302 |
+
"grad_norm": 0.3058314025402069,
|
| 303 |
+
"kl": 0.0005173740908503532,
|
| 304 |
+
"learning_rate": 5e-06,
|
| 305 |
+
"loss": 0.0001,
|
| 306 |
+
"reward": 0.32421875,
|
| 307 |
+
"reward_std": 0.2936626195907593,
|
| 308 |
+
"rewards/acc_reward_func": 0.32421875,
|
| 309 |
+
"step": 125
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"completion_length": 164.896875,
|
| 313 |
+
"epoch": 0.5220883534136547,
|
| 314 |
+
"grad_norm": 0.3665063977241516,
|
| 315 |
+
"kl": 0.0005303317215293646,
|
| 316 |
+
"learning_rate": 5e-06,
|
| 317 |
+
"loss": 0.0001,
|
| 318 |
+
"reward": 0.346875,
|
| 319 |
+
"reward_std": 0.28919236958026884,
|
| 320 |
+
"rewards/acc_reward_func": 0.346875,
|
| 321 |
+
"step": 130
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"completion_length": 169.23515625,
|
| 325 |
+
"epoch": 0.5421686746987951,
|
| 326 |
+
"grad_norm": 0.38457340002059937,
|
| 327 |
+
"kl": 0.0005899963027331979,
|
| 328 |
+
"learning_rate": 5e-06,
|
| 329 |
+
"loss": 0.0001,
|
| 330 |
+
"reward": 0.3453125,
|
| 331 |
+
"reward_std": 0.25805322229862215,
|
| 332 |
+
"rewards/acc_reward_func": 0.3453125,
|
| 333 |
+
"step": 135
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"completion_length": 164.45390625,
|
| 337 |
+
"epoch": 0.5622489959839357,
|
| 338 |
+
"grad_norm": 0.2263726145029068,
|
| 339 |
+
"kl": 0.0005681793205440045,
|
| 340 |
+
"learning_rate": 5e-06,
|
| 341 |
+
"loss": 0.0001,
|
| 342 |
+
"reward": 0.27890625,
|
| 343 |
+
"reward_std": 0.23708621561527252,
|
| 344 |
+
"rewards/acc_reward_func": 0.27890625,
|
| 345 |
+
"step": 140
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"completion_length": 167.23046875,
|
| 349 |
+
"epoch": 0.5823293172690763,
|
| 350 |
+
"grad_norm": 0.28835389018058777,
|
| 351 |
+
"kl": 0.0005865491810254752,
|
| 352 |
+
"learning_rate": 5e-06,
|
| 353 |
+
"loss": 0.0001,
|
| 354 |
+
"reward": 0.340625,
|
| 355 |
+
"reward_std": 0.2691764384508133,
|
| 356 |
+
"rewards/acc_reward_func": 0.340625,
|
| 357 |
+
"step": 145
|
| 358 |
+
},
|
| 359 |
+
{
|
| 360 |
+
"completion_length": 160.4046875,
|
| 361 |
+
"epoch": 0.6024096385542169,
|
| 362 |
+
"grad_norm": 0.2847937345504761,
|
| 363 |
+
"kl": 0.0006729792105033994,
|
| 364 |
+
"learning_rate": 5e-06,
|
| 365 |
+
"loss": 0.0001,
|
| 366 |
+
"reward": 0.3046875,
|
| 367 |
+
"reward_std": 0.2551824957132339,
|
| 368 |
+
"rewards/acc_reward_func": 0.3046875,
|
| 369 |
+
"step": 150
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"completion_length": 160.740625,
|
| 373 |
+
"epoch": 0.6224899598393574,
|
| 374 |
+
"grad_norm": 0.41062450408935547,
|
| 375 |
+
"kl": 0.0006228600163012743,
|
| 376 |
+
"learning_rate": 5e-06,
|
| 377 |
+
"loss": 0.0001,
|
| 378 |
+
"reward": 0.3265625,
|
| 379 |
+
"reward_std": 0.24656105935573577,
|
| 380 |
+
"rewards/acc_reward_func": 0.3265625,
|
| 381 |
+
"step": 155
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"completion_length": 168.340625,
|
| 385 |
+
"epoch": 0.642570281124498,
|
| 386 |
+
"grad_norm": 0.33770281076431274,
|
| 387 |
+
"kl": 0.0009516201331280172,
|
| 388 |
+
"learning_rate": 5e-06,
|
| 389 |
+
"loss": 0.0001,
|
| 390 |
+
"reward": 0.3609375,
|
| 391 |
+
"reward_std": 0.2608398377895355,
|
| 392 |
+
"rewards/acc_reward_func": 0.3609375,
|
| 393 |
+
"step": 160
|
| 394 |
+
},
|
| 395 |
+
{
|
| 396 |
+
"completion_length": 161.2953125,
|
| 397 |
+
"epoch": 0.6626506024096386,
|
| 398 |
+
"grad_norm": 0.3424857556819916,
|
| 399 |
+
"kl": 0.0007472435943782329,
|
| 400 |
+
"learning_rate": 5e-06,
|
| 401 |
+
"loss": 0.0001,
|
| 402 |
+
"reward": 0.3734375,
|
| 403 |
+
"reward_std": 0.292040029168129,
|
| 404 |
+
"rewards/acc_reward_func": 0.3734375,
|
| 405 |
+
"step": 165
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"completion_length": 173.7203125,
|
| 409 |
+
"epoch": 0.6827309236947792,
|
| 410 |
+
"grad_norm": 0.24203689396381378,
|
| 411 |
+
"kl": 0.0007647084421478212,
|
| 412 |
+
"learning_rate": 5e-06,
|
| 413 |
+
"loss": 0.0001,
|
| 414 |
+
"reward": 0.28203125,
|
| 415 |
+
"reward_std": 0.2562589019536972,
|
| 416 |
+
"rewards/acc_reward_func": 0.28203125,
|
| 417 |
+
"step": 170
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"completion_length": 167.025,
|
| 421 |
+
"epoch": 0.7028112449799196,
|
| 422 |
+
"grad_norm": 0.34411001205444336,
|
| 423 |
+
"kl": 0.000770233990624547,
|
| 424 |
+
"learning_rate": 5e-06,
|
| 425 |
+
"loss": 0.0001,
|
| 426 |
+
"reward": 0.32578125,
|
| 427 |
+
"reward_std": 0.25892365276813506,
|
| 428 |
+
"rewards/acc_reward_func": 0.32578125,
|
| 429 |
+
"step": 175
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"completion_length": 172.24453125,
|
| 433 |
+
"epoch": 0.7228915662650602,
|
| 434 |
+
"grad_norm": 0.29481959342956543,
|
| 435 |
+
"kl": 0.0009015992400236428,
|
| 436 |
+
"learning_rate": 5e-06,
|
| 437 |
+
"loss": 0.0001,
|
| 438 |
+
"reward": 0.340625,
|
| 439 |
+
"reward_std": 0.25950155556201937,
|
| 440 |
+
"rewards/acc_reward_func": 0.340625,
|
| 441 |
+
"step": 180
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"completion_length": 166.625,
|
| 445 |
+
"epoch": 0.7429718875502008,
|
| 446 |
+
"grad_norm": 0.2277025729417801,
|
| 447 |
+
"kl": 0.00077429274097085,
|
| 448 |
+
"learning_rate": 5e-06,
|
| 449 |
+
"loss": 0.0001,
|
| 450 |
+
"reward": 0.334375,
|
| 451 |
+
"reward_std": 0.2525084614753723,
|
| 452 |
+
"rewards/acc_reward_func": 0.334375,
|
| 453 |
+
"step": 185
|
| 454 |
+
},
|
| 455 |
+
{
|
| 456 |
+
"completion_length": 159.42734375,
|
| 457 |
+
"epoch": 0.7630522088353414,
|
| 458 |
+
"grad_norm": 0.32006603479385376,
|
| 459 |
+
"kl": 0.0008042196277529001,
|
| 460 |
+
"learning_rate": 5e-06,
|
| 461 |
+
"loss": 0.0001,
|
| 462 |
+
"reward": 0.353125,
|
| 463 |
+
"reward_std": 0.28777270913124087,
|
| 464 |
+
"rewards/acc_reward_func": 0.353125,
|
| 465 |
+
"step": 190
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"completion_length": 164.315625,
|
| 469 |
+
"epoch": 0.7831325301204819,
|
| 470 |
+
"grad_norm": 0.42280659079551697,
|
| 471 |
+
"kl": 0.0008120649959892035,
|
| 472 |
+
"learning_rate": 5e-06,
|
| 473 |
+
"loss": 0.0001,
|
| 474 |
+
"reward": 0.34140625,
|
| 475 |
+
"reward_std": 0.2515306770801544,
|
| 476 |
+
"rewards/acc_reward_func": 0.34140625,
|
| 477 |
+
"step": 195
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"completion_length": 163.465625,
|
| 481 |
+
"epoch": 0.8032128514056225,
|
| 482 |
+
"grad_norm": 0.3453792333602905,
|
| 483 |
+
"kl": 0.0007391226128675044,
|
| 484 |
+
"learning_rate": 5e-06,
|
| 485 |
+
"loss": 0.0001,
|
| 486 |
+
"reward": 0.33984375,
|
| 487 |
+
"reward_std": 0.2593150854110718,
|
| 488 |
+
"rewards/acc_reward_func": 0.33984375,
|
| 489 |
+
"step": 200
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"completion_length": 160.57109375,
|
| 493 |
+
"epoch": 0.8232931726907631,
|
| 494 |
+
"grad_norm": 0.22764600813388824,
|
| 495 |
+
"kl": 0.0009074539528228342,
|
| 496 |
+
"learning_rate": 5e-06,
|
| 497 |
+
"loss": 0.0001,
|
| 498 |
+
"reward": 0.33359375,
|
| 499 |
+
"reward_std": 0.24634140729904175,
|
| 500 |
+
"rewards/acc_reward_func": 0.33359375,
|
| 501 |
+
"step": 205
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"completion_length": 173.69453125,
|
| 505 |
+
"epoch": 0.8433734939759037,
|
| 506 |
+
"grad_norm": 0.3373042941093445,
|
| 507 |
+
"kl": 0.0008413837174884975,
|
| 508 |
+
"learning_rate": 5e-06,
|
| 509 |
+
"loss": 0.0001,
|
| 510 |
+
"reward": 0.32734375,
|
| 511 |
+
"reward_std": 0.2701482236385345,
|
| 512 |
+
"rewards/acc_reward_func": 0.32734375,
|
| 513 |
+
"step": 210
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"completion_length": 163.95546875,
|
| 517 |
+
"epoch": 0.8634538152610441,
|
| 518 |
+
"grad_norm": 0.43492230772972107,
|
| 519 |
+
"kl": 0.0008612593519501388,
|
| 520 |
+
"learning_rate": 5e-06,
|
| 521 |
+
"loss": 0.0001,
|
| 522 |
+
"reward": 0.35546875,
|
| 523 |
+
"reward_std": 0.25468774139881134,
|
| 524 |
+
"rewards/acc_reward_func": 0.35546875,
|
| 525 |
+
"step": 215
|
| 526 |
+
},
|
| 527 |
+
{
|
| 528 |
+
"completion_length": 165.01953125,
|
| 529 |
+
"epoch": 0.8835341365461847,
|
| 530 |
+
"grad_norm": 0.47059279680252075,
|
| 531 |
+
"kl": 0.0007435820298269391,
|
| 532 |
+
"learning_rate": 5e-06,
|
| 533 |
+
"loss": 0.0001,
|
| 534 |
+
"reward": 0.3390625,
|
| 535 |
+
"reward_std": 0.26499341428279877,
|
| 536 |
+
"rewards/acc_reward_func": 0.3390625,
|
| 537 |
+
"step": 220
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"completion_length": 169.05078125,
|
| 541 |
+
"epoch": 0.9036144578313253,
|
| 542 |
+
"grad_norm": 0.22417674958705902,
|
| 543 |
+
"kl": 0.0007649007253348828,
|
| 544 |
+
"learning_rate": 5e-06,
|
| 545 |
+
"loss": 0.0001,
|
| 546 |
+
"reward": 0.284375,
|
| 547 |
+
"reward_std": 0.2533974200487137,
|
| 548 |
+
"rewards/acc_reward_func": 0.284375,
|
| 549 |
+
"step": 225
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"completion_length": 173.67109375,
|
| 553 |
+
"epoch": 0.9236947791164659,
|
| 554 |
+
"grad_norm": 0.2978118658065796,
|
| 555 |
+
"kl": 0.0007940458599478006,
|
| 556 |
+
"learning_rate": 5e-06,
|
| 557 |
+
"loss": 0.0001,
|
| 558 |
+
"reward": 0.2984375,
|
| 559 |
+
"reward_std": 0.24650274217128754,
|
| 560 |
+
"rewards/acc_reward_func": 0.2984375,
|
| 561 |
+
"step": 230
|
| 562 |
+
},
|
| 563 |
+
{
|
| 564 |
+
"completion_length": 167.99375,
|
| 565 |
+
"epoch": 0.9437751004016064,
|
| 566 |
+
"grad_norm": 0.2792234420776367,
|
| 567 |
+
"kl": 0.0010552789666689933,
|
| 568 |
+
"learning_rate": 5e-06,
|
| 569 |
+
"loss": 0.0001,
|
| 570 |
+
"reward": 0.35,
|
| 571 |
+
"reward_std": 0.26749635934829713,
|
| 572 |
+
"rewards/acc_reward_func": 0.35,
|
| 573 |
+
"step": 235
|
| 574 |
+
},
|
| 575 |
+
{
|
| 576 |
+
"completion_length": 171.94609375,
|
| 577 |
+
"epoch": 0.963855421686747,
|
| 578 |
+
"grad_norm": 0.2678660452365875,
|
| 579 |
+
"kl": 0.0006605981849133969,
|
| 580 |
+
"learning_rate": 5e-06,
|
| 581 |
+
"loss": 0.0001,
|
| 582 |
+
"reward": 0.31875,
|
| 583 |
+
"reward_std": 0.2612347215414047,
|
| 584 |
+
"rewards/acc_reward_func": 0.31875,
|
| 585 |
+
"step": 240
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"completion_length": 160.3421875,
|
| 589 |
+
"epoch": 0.9839357429718876,
|
| 590 |
+
"grad_norm": 0.6757539510726929,
|
| 591 |
+
"kl": 0.0006685945438221097,
|
| 592 |
+
"learning_rate": 5e-06,
|
| 593 |
+
"loss": 0.0001,
|
| 594 |
+
"reward": 0.33984375,
|
| 595 |
+
"reward_std": 0.2489775002002716,
|
| 596 |
+
"rewards/acc_reward_func": 0.33984375,
|
| 597 |
+
"step": 245
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"completion_length": 159.35982360839844,
|
| 601 |
+
"epoch": 1.0040160642570282,
|
| 602 |
+
"grad_norm": 0.4328139126300812,
|
| 603 |
+
"kl": 0.0009481518063694239,
|
| 604 |
+
"learning_rate": 5e-06,
|
| 605 |
+
"loss": 0.0001,
|
| 606 |
+
"reward": 0.3328125,
|
| 607 |
+
"reward_std": 0.27164973616600036,
|
| 608 |
+
"rewards/acc_reward_func": 0.3328125,
|
| 609 |
+
"step": 250
|
| 610 |
+
},
|
| 611 |
+
{
|
| 612 |
+
"completion_length": 170.5609375,
|
| 613 |
+
"epoch": 1.0240963855421688,
|
| 614 |
+
"grad_norm": 0.40751489996910095,
|
| 615 |
+
"kl": 0.0014292935142293573,
|
| 616 |
+
"learning_rate": 5e-06,
|
| 617 |
+
"loss": 0.0001,
|
| 618 |
+
"reward": 0.34296875,
|
| 619 |
+
"reward_std": 0.26451934576034547,
|
| 620 |
+
"rewards/acc_reward_func": 0.34296875,
|
| 621 |
+
"step": 255
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"completion_length": 158.06484375,
|
| 625 |
+
"epoch": 1.0441767068273093,
|
| 626 |
+
"grad_norm": 0.32165759801864624,
|
| 627 |
+
"kl": 0.0023219846189022064,
|
| 628 |
+
"learning_rate": 5e-06,
|
| 629 |
+
"loss": 0.0002,
|
| 630 |
+
"reward": 0.31796875,
|
| 631 |
+
"reward_std": 0.26662840247154235,
|
| 632 |
+
"rewards/acc_reward_func": 0.31796875,
|
| 633 |
+
"step": 260
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"completion_length": 162.784375,
|
| 637 |
+
"epoch": 1.0642570281124497,
|
| 638 |
+
"grad_norm": 0.4819350838661194,
|
| 639 |
+
"kl": 0.0021671449765563013,
|
| 640 |
+
"learning_rate": 5e-06,
|
| 641 |
+
"loss": 0.0002,
|
| 642 |
+
"reward": 0.3671875,
|
| 643 |
+
"reward_std": 0.2797468721866608,
|
| 644 |
+
"rewards/acc_reward_func": 0.3671875,
|
| 645 |
+
"step": 265
|
| 646 |
+
},
|
| 647 |
+
{
|
| 648 |
+
"completion_length": 174.1890625,
|
| 649 |
+
"epoch": 1.0843373493975903,
|
| 650 |
+
"grad_norm": 0.2753521502017975,
|
| 651 |
+
"kl": 0.0023205589037388562,
|
| 652 |
+
"learning_rate": 5e-06,
|
| 653 |
+
"loss": 0.0002,
|
| 654 |
+
"reward": 0.33984375,
|
| 655 |
+
"reward_std": 0.2735643357038498,
|
| 656 |
+
"rewards/acc_reward_func": 0.33984375,
|
| 657 |
+
"step": 270
|
| 658 |
+
},
|
| 659 |
+
{
|
| 660 |
+
"completion_length": 173.61640625,
|
| 661 |
+
"epoch": 1.104417670682731,
|
| 662 |
+
"grad_norm": 0.2579421103000641,
|
| 663 |
+
"kl": 0.0014010543003678323,
|
| 664 |
+
"learning_rate": 5e-06,
|
| 665 |
+
"loss": 0.0001,
|
| 666 |
+
"reward": 0.2578125,
|
| 667 |
+
"reward_std": 0.2305402159690857,
|
| 668 |
+
"rewards/acc_reward_func": 0.2578125,
|
| 669 |
+
"step": 275
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"completion_length": 168.61640625,
|
| 673 |
+
"epoch": 1.1244979919678715,
|
| 674 |
+
"grad_norm": 0.3602357506752014,
|
| 675 |
+
"kl": 0.0012595997890457512,
|
| 676 |
+
"learning_rate": 5e-06,
|
| 677 |
+
"loss": 0.0001,
|
| 678 |
+
"reward": 0.359375,
|
| 679 |
+
"reward_std": 0.3013946235179901,
|
| 680 |
+
"rewards/acc_reward_func": 0.359375,
|
| 681 |
+
"step": 280
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"completion_length": 171.4640625,
|
| 685 |
+
"epoch": 1.144578313253012,
|
| 686 |
+
"grad_norm": 0.2711893320083618,
|
| 687 |
+
"kl": 0.0009470222401432693,
|
| 688 |
+
"learning_rate": 5e-06,
|
| 689 |
+
"loss": 0.0001,
|
| 690 |
+
"reward": 0.35390625,
|
| 691 |
+
"reward_std": 0.27838287949562074,
|
| 692 |
+
"rewards/acc_reward_func": 0.35390625,
|
| 693 |
+
"step": 285
|
| 694 |
+
},
|
| 695 |
+
{
|
| 696 |
+
"completion_length": 172.66015625,
|
| 697 |
+
"epoch": 1.1646586345381527,
|
| 698 |
+
"grad_norm": 0.2950332760810852,
|
| 699 |
+
"kl": 0.0009261242463253439,
|
| 700 |
+
"learning_rate": 5e-06,
|
| 701 |
+
"loss": 0.0001,
|
| 702 |
+
"reward": 0.33359375,
|
| 703 |
+
"reward_std": 0.28246039152145386,
|
| 704 |
+
"rewards/acc_reward_func": 0.33359375,
|
| 705 |
+
"step": 290
|
| 706 |
+
},
|
| 707 |
+
{
|
| 708 |
+
"completion_length": 174.0375,
|
| 709 |
+
"epoch": 1.1847389558232932,
|
| 710 |
+
"grad_norm": 0.42756450176239014,
|
| 711 |
+
"kl": 0.0008738423697650432,
|
| 712 |
+
"learning_rate": 5e-06,
|
| 713 |
+
"loss": 0.0001,
|
| 714 |
+
"reward": 0.3171875,
|
| 715 |
+
"reward_std": 0.268758499622345,
|
| 716 |
+
"rewards/acc_reward_func": 0.3171875,
|
| 717 |
+
"step": 295
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"completion_length": 163.82265625,
|
| 721 |
+
"epoch": 1.2048192771084336,
|
| 722 |
+
"grad_norm": 0.4912799000740051,
|
| 723 |
+
"kl": 0.001153232657816261,
|
| 724 |
+
"learning_rate": 5e-06,
|
| 725 |
+
"loss": 0.0001,
|
| 726 |
+
"reward": 0.33359375,
|
| 727 |
+
"reward_std": 0.2731469988822937,
|
| 728 |
+
"rewards/acc_reward_func": 0.33359375,
|
| 729 |
+
"step": 300
|
| 730 |
+
},
|
| 731 |
+
{
|
| 732 |
+
"completion_length": 158.7875,
|
| 733 |
+
"epoch": 1.2248995983935742,
|
| 734 |
+
"grad_norm": 0.30150240659713745,
|
| 735 |
+
"kl": 0.0008952352683991194,
|
| 736 |
+
"learning_rate": 5e-06,
|
| 737 |
+
"loss": 0.0001,
|
| 738 |
+
"reward": 0.3546875,
|
| 739 |
+
"reward_std": 0.27324608862400057,
|
| 740 |
+
"rewards/acc_reward_func": 0.3546875,
|
| 741 |
+
"step": 305
|
| 742 |
+
},
|
| 743 |
+
{
|
| 744 |
+
"completion_length": 176.72421875,
|
| 745 |
+
"epoch": 1.2449799196787148,
|
| 746 |
+
"grad_norm": 0.35327720642089844,
|
| 747 |
+
"kl": 0.0011502272100187838,
|
| 748 |
+
"learning_rate": 5e-06,
|
| 749 |
+
"loss": 0.0001,
|
| 750 |
+
"reward": 0.35625,
|
| 751 |
+
"reward_std": 0.27567465901374816,
|
| 752 |
+
"rewards/acc_reward_func": 0.35625,
|
| 753 |
+
"step": 310
|
| 754 |
+
},
|
| 755 |
+
{
|
| 756 |
+
"completion_length": 166.778125,
|
| 757 |
+
"epoch": 1.2650602409638554,
|
| 758 |
+
"grad_norm": 0.3498896360397339,
|
| 759 |
+
"kl": 0.0016141865635290742,
|
| 760 |
+
"learning_rate": 5e-06,
|
| 761 |
+
"loss": 0.0002,
|
| 762 |
+
"reward": 0.30703125,
|
| 763 |
+
"reward_std": 0.261915448307991,
|
| 764 |
+
"rewards/acc_reward_func": 0.30703125,
|
| 765 |
+
"step": 315
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"completion_length": 178.128125,
|
| 769 |
+
"epoch": 1.285140562248996,
|
| 770 |
+
"grad_norm": 0.297959566116333,
|
| 771 |
+
"kl": 0.00151687542675063,
|
| 772 |
+
"learning_rate": 5e-06,
|
| 773 |
+
"loss": 0.0002,
|
| 774 |
+
"reward": 0.32421875,
|
| 775 |
+
"reward_std": 0.27393734753131865,
|
| 776 |
+
"rewards/acc_reward_func": 0.32421875,
|
| 777 |
+
"step": 320
|
| 778 |
+
},
|
| 779 |
+
{
|
| 780 |
+
"completion_length": 170.6953125,
|
| 781 |
+
"epoch": 1.3052208835341366,
|
| 782 |
+
"grad_norm": 0.365997850894928,
|
| 783 |
+
"kl": 0.0010963335167616605,
|
| 784 |
+
"learning_rate": 5e-06,
|
| 785 |
+
"loss": 0.0001,
|
| 786 |
+
"reward": 0.315625,
|
| 787 |
+
"reward_std": 0.2619461864233017,
|
| 788 |
+
"rewards/acc_reward_func": 0.315625,
|
| 789 |
+
"step": 325
|
| 790 |
+
},
|
| 791 |
+
{
|
| 792 |
+
"completion_length": 166.228125,
|
| 793 |
+
"epoch": 1.3253012048192772,
|
| 794 |
+
"grad_norm": 0.36575648188591003,
|
| 795 |
+
"kl": 0.0010827683610841632,
|
| 796 |
+
"learning_rate": 5e-06,
|
| 797 |
+
"loss": 0.0001,
|
| 798 |
+
"reward": 0.31015625,
|
| 799 |
+
"reward_std": 0.26013057231903075,
|
| 800 |
+
"rewards/acc_reward_func": 0.31015625,
|
| 801 |
+
"step": 330
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"completion_length": 166.16484375,
|
| 805 |
+
"epoch": 1.3453815261044177,
|
| 806 |
+
"grad_norm": 0.29126739501953125,
|
| 807 |
+
"kl": 0.0009882883401587605,
|
| 808 |
+
"learning_rate": 5e-06,
|
| 809 |
+
"loss": 0.0001,
|
| 810 |
+
"reward": 0.309375,
|
| 811 |
+
"reward_std": 0.25576087832450867,
|
| 812 |
+
"rewards/acc_reward_func": 0.309375,
|
| 813 |
+
"step": 335
|
| 814 |
+
},
|
| 815 |
+
{
|
| 816 |
+
"completion_length": 173.39453125,
|
| 817 |
+
"epoch": 1.3654618473895583,
|
| 818 |
+
"grad_norm": 0.33560770750045776,
|
| 819 |
+
"kl": 0.0011842920910567045,
|
| 820 |
+
"learning_rate": 5e-06,
|
| 821 |
+
"loss": 0.0001,
|
| 822 |
+
"reward": 0.31953125,
|
| 823 |
+
"reward_std": 0.2704659789800644,
|
| 824 |
+
"rewards/acc_reward_func": 0.31953125,
|
| 825 |
+
"step": 340
|
| 826 |
+
},
|
| 827 |
+
{
|
| 828 |
+
"completion_length": 173.82578125,
|
| 829 |
+
"epoch": 1.3855421686746987,
|
| 830 |
+
"grad_norm": 0.2982430160045624,
|
| 831 |
+
"kl": 0.0015789813129231333,
|
| 832 |
+
"learning_rate": 5e-06,
|
| 833 |
+
"loss": 0.0002,
|
| 834 |
+
"reward": 0.34453125,
|
| 835 |
+
"reward_std": 0.3086081326007843,
|
| 836 |
+
"rewards/acc_reward_func": 0.34453125,
|
| 837 |
+
"step": 345
|
| 838 |
+
},
|
| 839 |
+
{
|
| 840 |
+
"completion_length": 162.96953125,
|
| 841 |
+
"epoch": 1.4056224899598393,
|
| 842 |
+
"grad_norm": 0.5575153827667236,
|
| 843 |
+
"kl": 0.001101066661067307,
|
| 844 |
+
"learning_rate": 5e-06,
|
| 845 |
+
"loss": 0.0001,
|
| 846 |
+
"reward": 0.3765625,
|
| 847 |
+
"reward_std": 0.29579696655273435,
|
| 848 |
+
"rewards/acc_reward_func": 0.3765625,
|
| 849 |
+
"step": 350
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"completion_length": 181.57578125,
|
| 853 |
+
"epoch": 1.4257028112449799,
|
| 854 |
+
"grad_norm": 0.5113179683685303,
|
| 855 |
+
"kl": 0.001085875742137432,
|
| 856 |
+
"learning_rate": 5e-06,
|
| 857 |
+
"loss": 0.0001,
|
| 858 |
+
"reward": 0.29609375,
|
| 859 |
+
"reward_std": 0.26413264572620393,
|
| 860 |
+
"rewards/acc_reward_func": 0.29609375,
|
| 861 |
+
"step": 355
|
| 862 |
+
},
|
| 863 |
+
{
|
| 864 |
+
"completion_length": 174.1046875,
|
| 865 |
+
"epoch": 1.4457831325301205,
|
| 866 |
+
"grad_norm": 0.47564586997032166,
|
| 867 |
+
"kl": 0.0010274604661390184,
|
| 868 |
+
"learning_rate": 5e-06,
|
| 869 |
+
"loss": 0.0001,
|
| 870 |
+
"reward": 0.32890625,
|
| 871 |
+
"reward_std": 0.2666553735733032,
|
| 872 |
+
"rewards/acc_reward_func": 0.32890625,
|
| 873 |
+
"step": 360
|
| 874 |
+
},
|
| 875 |
+
{
|
| 876 |
+
"completion_length": 164.609375,
|
| 877 |
+
"epoch": 1.465863453815261,
|
| 878 |
+
"grad_norm": 0.23254898190498352,
|
| 879 |
+
"kl": 0.0011269458453170955,
|
| 880 |
+
"learning_rate": 5e-06,
|
| 881 |
+
"loss": 0.0001,
|
| 882 |
+
"reward": 0.3921875,
|
| 883 |
+
"reward_std": 0.26207175552845,
|
| 884 |
+
"rewards/acc_reward_func": 0.3921875,
|
| 885 |
+
"step": 365
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"completion_length": 171.3859375,
|
| 889 |
+
"epoch": 1.4859437751004017,
|
| 890 |
+
"grad_norm": 0.2955392599105835,
|
| 891 |
+
"kl": 0.0013015888049267232,
|
| 892 |
+
"learning_rate": 5e-06,
|
| 893 |
+
"loss": 0.0001,
|
| 894 |
+
"reward": 0.3390625,
|
| 895 |
+
"reward_std": 0.2618942677974701,
|
| 896 |
+
"rewards/acc_reward_func": 0.3390625,
|
| 897 |
+
"step": 370
|
| 898 |
+
},
|
| 899 |
+
{
|
| 900 |
+
"completion_length": 170.51328125,
|
| 901 |
+
"epoch": 1.5060240963855422,
|
| 902 |
+
"grad_norm": 0.2367907017469406,
|
| 903 |
+
"kl": 0.0012052926933392883,
|
| 904 |
+
"learning_rate": 5e-06,
|
| 905 |
+
"loss": 0.0001,
|
| 906 |
+
"reward": 0.30859375,
|
| 907 |
+
"reward_std": 0.243873330950737,
|
| 908 |
+
"rewards/acc_reward_func": 0.30859375,
|
| 909 |
+
"step": 375
|
| 910 |
+
},
|
| 911 |
+
{
|
| 912 |
+
"completion_length": 171.328125,
|
| 913 |
+
"epoch": 1.5261044176706826,
|
| 914 |
+
"grad_norm": 0.4387330412864685,
|
| 915 |
+
"kl": 0.0011040043318644166,
|
| 916 |
+
"learning_rate": 5e-06,
|
| 917 |
+
"loss": 0.0001,
|
| 918 |
+
"reward": 0.34296875,
|
| 919 |
+
"reward_std": 0.27786171436309814,
|
| 920 |
+
"rewards/acc_reward_func": 0.34296875,
|
| 921 |
+
"step": 380
|
| 922 |
+
},
|
| 923 |
+
{
|
| 924 |
+
"completion_length": 167.015625,
|
| 925 |
+
"epoch": 1.5461847389558234,
|
| 926 |
+
"grad_norm": 0.2528681457042694,
|
| 927 |
+
"kl": 0.001643406949006021,
|
| 928 |
+
"learning_rate": 5e-06,
|
| 929 |
+
"loss": 0.0002,
|
| 930 |
+
"reward": 0.36484375,
|
| 931 |
+
"reward_std": 0.3019732892513275,
|
| 932 |
+
"rewards/acc_reward_func": 0.36484375,
|
| 933 |
+
"step": 385
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"completion_length": 150.4078125,
|
| 937 |
+
"epoch": 1.5662650602409638,
|
| 938 |
+
"grad_norm": 0.35000428557395935,
|
| 939 |
+
"kl": 0.0013000907842069865,
|
| 940 |
+
"learning_rate": 5e-06,
|
| 941 |
+
"loss": 0.0001,
|
| 942 |
+
"reward": 0.378125,
|
| 943 |
+
"reward_std": 0.28335249423980713,
|
| 944 |
+
"rewards/acc_reward_func": 0.378125,
|
| 945 |
+
"step": 390
|
| 946 |
+
},
|
| 947 |
+
{
|
| 948 |
+
"completion_length": 158.709375,
|
| 949 |
+
"epoch": 1.5863453815261044,
|
| 950 |
+
"grad_norm": 0.259275883436203,
|
| 951 |
+
"kl": 0.0016851373482495546,
|
| 952 |
+
"learning_rate": 5e-06,
|
| 953 |
+
"loss": 0.0002,
|
| 954 |
+
"reward": 0.31953125,
|
| 955 |
+
"reward_std": 0.28603203892707824,
|
| 956 |
+
"rewards/acc_reward_func": 0.31953125,
|
| 957 |
+
"step": 395
|
| 958 |
+
},
|
| 959 |
+
{
|
| 960 |
+
"completion_length": 160.56796875,
|
| 961 |
+
"epoch": 1.606425702811245,
|
| 962 |
+
"grad_norm": 0.31947171688079834,
|
| 963 |
+
"kl": 0.001047816756181419,
|
| 964 |
+
"learning_rate": 5e-06,
|
| 965 |
+
"loss": 0.0001,
|
| 966 |
+
"reward": 0.3359375,
|
| 967 |
+
"reward_std": 0.27272760272026064,
|
| 968 |
+
"rewards/acc_reward_func": 0.3359375,
|
| 969 |
+
"step": 400
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"completion_length": 161.90234375,
|
| 973 |
+
"epoch": 1.6265060240963856,
|
| 974 |
+
"grad_norm": 0.34025779366493225,
|
| 975 |
+
"kl": 0.0011820008745416998,
|
| 976 |
+
"learning_rate": 5e-06,
|
| 977 |
+
"loss": 0.0001,
|
| 978 |
+
"reward": 0.33671875,
|
| 979 |
+
"reward_std": 0.2665954947471619,
|
| 980 |
+
"rewards/acc_reward_func": 0.33671875,
|
| 981 |
+
"step": 405
|
| 982 |
+
},
|
| 983 |
+
{
|
| 984 |
+
"completion_length": 167.6984375,
|
| 985 |
+
"epoch": 1.6465863453815262,
|
| 986 |
+
"grad_norm": 0.2868824005126953,
|
| 987 |
+
"kl": 0.0010516393929719924,
|
| 988 |
+
"learning_rate": 5e-06,
|
| 989 |
+
"loss": 0.0001,
|
| 990 |
+
"reward": 0.3578125,
|
| 991 |
+
"reward_std": 0.2612337410449982,
|
| 992 |
+
"rewards/acc_reward_func": 0.3578125,
|
| 993 |
+
"step": 410
|
| 994 |
+
},
|
| 995 |
+
{
|
| 996 |
+
"completion_length": 166.8375,
|
| 997 |
+
"epoch": 1.6666666666666665,
|
| 998 |
+
"grad_norm": 0.30937594175338745,
|
| 999 |
+
"kl": 0.0011767351999878884,
|
| 1000 |
+
"learning_rate": 5e-06,
|
| 1001 |
+
"loss": 0.0001,
|
| 1002 |
+
"reward": 0.30859375,
|
| 1003 |
+
"reward_std": 0.2570806533098221,
|
| 1004 |
+
"rewards/acc_reward_func": 0.30859375,
|
| 1005 |
+
"step": 415
|
| 1006 |
+
},
|
| 1007 |
+
{
|
| 1008 |
+
"completion_length": 170.19765625,
|
| 1009 |
+
"epoch": 1.6867469879518073,
|
| 1010 |
+
"grad_norm": 0.25181344151496887,
|
| 1011 |
+
"kl": 0.0010755043127574026,
|
| 1012 |
+
"learning_rate": 5e-06,
|
| 1013 |
+
"loss": 0.0001,
|
| 1014 |
+
"reward": 0.33125,
|
| 1015 |
+
"reward_std": 0.22994134724140167,
|
| 1016 |
+
"rewards/acc_reward_func": 0.33125,
|
| 1017 |
+
"step": 420
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"completion_length": 165.94609375,
|
| 1021 |
+
"epoch": 1.7068273092369477,
|
| 1022 |
+
"grad_norm": 0.396030992269516,
|
| 1023 |
+
"kl": 0.0012019906658679246,
|
| 1024 |
+
"learning_rate": 5e-06,
|
| 1025 |
+
"loss": 0.0001,
|
| 1026 |
+
"reward": 0.315625,
|
| 1027 |
+
"reward_std": 0.27038955092430117,
|
| 1028 |
+
"rewards/acc_reward_func": 0.315625,
|
| 1029 |
+
"step": 425
|
| 1030 |
+
},
|
| 1031 |
+
{
|
| 1032 |
+
"completion_length": 167.24140625,
|
| 1033 |
+
"epoch": 1.7269076305220885,
|
| 1034 |
+
"grad_norm": 0.3269684314727783,
|
| 1035 |
+
"kl": 0.001297155674546957,
|
| 1036 |
+
"learning_rate": 5e-06,
|
| 1037 |
+
"loss": 0.0001,
|
| 1038 |
+
"reward": 0.32890625,
|
| 1039 |
+
"reward_std": 0.26612446308135984,
|
| 1040 |
+
"rewards/acc_reward_func": 0.32890625,
|
| 1041 |
+
"step": 430
|
| 1042 |
+
},
|
| 1043 |
+
{
|
| 1044 |
+
"completion_length": 155.46015625,
|
| 1045 |
+
"epoch": 1.7469879518072289,
|
| 1046 |
+
"grad_norm": 0.46771514415740967,
|
| 1047 |
+
"kl": 0.0014664881862699985,
|
| 1048 |
+
"learning_rate": 5e-06,
|
| 1049 |
+
"loss": 0.0001,
|
| 1050 |
+
"reward": 0.3234375,
|
| 1051 |
+
"reward_std": 0.24766394197940828,
|
| 1052 |
+
"rewards/acc_reward_func": 0.3234375,
|
| 1053 |
+
"step": 435
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"completion_length": 164.859375,
|
| 1057 |
+
"epoch": 1.7670682730923695,
|
| 1058 |
+
"grad_norm": 0.24466943740844727,
|
| 1059 |
+
"kl": 0.0012787181185558438,
|
| 1060 |
+
"learning_rate": 5e-06,
|
| 1061 |
+
"loss": 0.0001,
|
| 1062 |
+
"reward": 0.334375,
|
| 1063 |
+
"reward_std": 0.2795639634132385,
|
| 1064 |
+
"rewards/acc_reward_func": 0.334375,
|
| 1065 |
+
"step": 440
|
| 1066 |
+
},
|
| 1067 |
+
{
|
| 1068 |
+
"completion_length": 180.32578125,
|
| 1069 |
+
"epoch": 1.78714859437751,
|
| 1070 |
+
"grad_norm": 0.28328680992126465,
|
| 1071 |
+
"kl": 0.001317713246680796,
|
| 1072 |
+
"learning_rate": 5e-06,
|
| 1073 |
+
"loss": 0.0001,
|
| 1074 |
+
"reward": 0.29296875,
|
| 1075 |
+
"reward_std": 0.22741309106349944,
|
| 1076 |
+
"rewards/acc_reward_func": 0.29296875,
|
| 1077 |
+
"step": 445
|
| 1078 |
+
},
|
| 1079 |
+
{
|
| 1080 |
+
"completion_length": 163.96875,
|
| 1081 |
+
"epoch": 1.8072289156626506,
|
| 1082 |
+
"grad_norm": 0.27741825580596924,
|
| 1083 |
+
"kl": 0.0015525751281529666,
|
| 1084 |
+
"learning_rate": 5e-06,
|
| 1085 |
+
"loss": 0.0002,
|
| 1086 |
+
"reward": 0.32421875,
|
| 1087 |
+
"reward_std": 0.28004167675971986,
|
| 1088 |
+
"rewards/acc_reward_func": 0.32421875,
|
| 1089 |
+
"step": 450
|
| 1090 |
+
},
|
| 1091 |
+
{
|
| 1092 |
+
"completion_length": 165.8140625,
|
| 1093 |
+
"epoch": 1.8273092369477912,
|
| 1094 |
+
"grad_norm": 0.2740982472896576,
|
| 1095 |
+
"kl": 0.0013578152284026146,
|
| 1096 |
+
"learning_rate": 5e-06,
|
| 1097 |
+
"loss": 0.0001,
|
| 1098 |
+
"reward": 0.33203125,
|
| 1099 |
+
"reward_std": 0.2814855635166168,
|
| 1100 |
+
"rewards/acc_reward_func": 0.33203125,
|
| 1101 |
+
"step": 455
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"completion_length": 164.31171875,
|
| 1105 |
+
"epoch": 1.8473895582329316,
|
| 1106 |
+
"grad_norm": 0.4178365468978882,
|
| 1107 |
+
"kl": 0.0013977297581732272,
|
| 1108 |
+
"learning_rate": 5e-06,
|
| 1109 |
+
"loss": 0.0001,
|
| 1110 |
+
"reward": 0.36796875,
|
| 1111 |
+
"reward_std": 0.2695493370294571,
|
| 1112 |
+
"rewards/acc_reward_func": 0.36796875,
|
| 1113 |
+
"step": 460
|
| 1114 |
+
},
|
| 1115 |
+
{
|
| 1116 |
+
"completion_length": 176.71953125,
|
| 1117 |
+
"epoch": 1.8674698795180724,
|
| 1118 |
+
"grad_norm": 0.2551046311855316,
|
| 1119 |
+
"kl": 0.0013892159098759294,
|
| 1120 |
+
"learning_rate": 5e-06,
|
| 1121 |
+
"loss": 0.0001,
|
| 1122 |
+
"reward": 0.26328125,
|
| 1123 |
+
"reward_std": 0.24187707006931305,
|
| 1124 |
+
"rewards/acc_reward_func": 0.26328125,
|
| 1125 |
+
"step": 465
|
| 1126 |
+
},
|
| 1127 |
+
{
|
| 1128 |
+
"completion_length": 162.015625,
|
| 1129 |
+
"epoch": 1.8875502008032128,
|
| 1130 |
+
"grad_norm": 0.23321259021759033,
|
| 1131 |
+
"kl": 0.0014218664728105068,
|
| 1132 |
+
"learning_rate": 5e-06,
|
| 1133 |
+
"loss": 0.0001,
|
| 1134 |
+
"reward": 0.31328125,
|
| 1135 |
+
"reward_std": 0.24751116037368776,
|
| 1136 |
+
"rewards/acc_reward_func": 0.31328125,
|
| 1137 |
+
"step": 470
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"completion_length": 181.08125,
|
| 1141 |
+
"epoch": 1.9076305220883534,
|
| 1142 |
+
"grad_norm": 0.26958364248275757,
|
| 1143 |
+
"kl": 0.0012838932918384672,
|
| 1144 |
+
"learning_rate": 5e-06,
|
| 1145 |
+
"loss": 0.0001,
|
| 1146 |
+
"reward": 0.31328125,
|
| 1147 |
+
"reward_std": 0.28040671050548555,
|
| 1148 |
+
"rewards/acc_reward_func": 0.31328125,
|
| 1149 |
+
"step": 475
|
| 1150 |
+
},
|
| 1151 |
+
{
|
| 1152 |
+
"completion_length": 162.40390625,
|
| 1153 |
+
"epoch": 1.927710843373494,
|
| 1154 |
+
"grad_norm": 0.27885714173316956,
|
| 1155 |
+
"kl": 0.0011875152122229338,
|
| 1156 |
+
"learning_rate": 5e-06,
|
| 1157 |
+
"loss": 0.0001,
|
| 1158 |
+
"reward": 0.36484375,
|
| 1159 |
+
"reward_std": 0.26504404842853546,
|
| 1160 |
+
"rewards/acc_reward_func": 0.36484375,
|
| 1161 |
+
"step": 480
|
| 1162 |
+
},
|
| 1163 |
+
{
|
| 1164 |
+
"completion_length": 158.46640625,
|
| 1165 |
+
"epoch": 1.9477911646586346,
|
| 1166 |
+
"grad_norm": 0.27203765511512756,
|
| 1167 |
+
"kl": 0.0011519475607201456,
|
| 1168 |
+
"learning_rate": 5e-06,
|
| 1169 |
+
"loss": 0.0001,
|
| 1170 |
+
"reward": 0.35390625,
|
| 1171 |
+
"reward_std": 0.25064152777194976,
|
| 1172 |
+
"rewards/acc_reward_func": 0.35390625,
|
| 1173 |
+
"step": 485
|
| 1174 |
+
},
|
| 1175 |
+
{
|
| 1176 |
+
"completion_length": 163.49921875,
|
| 1177 |
+
"epoch": 1.9678714859437751,
|
| 1178 |
+
"grad_norm": 0.38953107595443726,
|
| 1179 |
+
"kl": 0.002528494060970843,
|
| 1180 |
+
"learning_rate": 5e-06,
|
| 1181 |
+
"loss": 0.0003,
|
| 1182 |
+
"reward": 0.33125,
|
| 1183 |
+
"reward_std": 0.265842866897583,
|
| 1184 |
+
"rewards/acc_reward_func": 0.33125,
|
| 1185 |
+
"step": 490
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"completion_length": 160.84140625,
|
| 1189 |
+
"epoch": 1.9879518072289155,
|
| 1190 |
+
"grad_norm": 0.268568754196167,
|
| 1191 |
+
"kl": 0.0014860291033983231,
|
| 1192 |
+
"learning_rate": 5e-06,
|
| 1193 |
+
"loss": 0.0001,
|
| 1194 |
+
"reward": 0.31953125,
|
| 1195 |
+
"reward_std": 0.25731430053710935,
|
| 1196 |
+
"rewards/acc_reward_func": 0.31953125,
|
| 1197 |
+
"step": 495
|
| 1198 |
+
},
|
| 1199 |
+
{
|
| 1200 |
+
"completion_length": 179.959375,
|
| 1201 |
+
"epoch": 2.0080321285140563,
|
| 1202 |
+
"grad_norm": 0.3279021680355072,
|
| 1203 |
+
"kl": 0.0021104462211951613,
|
| 1204 |
+
"learning_rate": 5e-06,
|
| 1205 |
+
"loss": 0.0002,
|
| 1206 |
+
"reward": 0.34140625,
|
| 1207 |
+
"reward_std": 0.2936858534812927,
|
| 1208 |
+
"rewards/acc_reward_func": 0.34140625,
|
| 1209 |
+
"step": 500
|
| 1210 |
+
},
|
| 1211 |
+
{
|
| 1212 |
+
"completion_length": 172.03984375,
|
| 1213 |
+
"epoch": 2.0281124497991967,
|
| 1214 |
+
"grad_norm": 0.30037155747413635,
|
| 1215 |
+
"kl": 0.001345854508690536,
|
| 1216 |
+
"learning_rate": 5e-06,
|
| 1217 |
+
"loss": 0.0001,
|
| 1218 |
+
"reward": 0.2671875,
|
| 1219 |
+
"reward_std": 0.23858949542045593,
|
| 1220 |
+
"rewards/acc_reward_func": 0.2671875,
|
| 1221 |
+
"step": 505
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"completion_length": 166.13359375,
|
| 1225 |
+
"epoch": 2.0481927710843375,
|
| 1226 |
+
"grad_norm": 0.24528227746486664,
|
| 1227 |
+
"kl": 0.0013967655366286635,
|
| 1228 |
+
"learning_rate": 5e-06,
|
| 1229 |
+
"loss": 0.0001,
|
| 1230 |
+
"reward": 0.28203125,
|
| 1231 |
+
"reward_std": 0.2324840843677521,
|
| 1232 |
+
"rewards/acc_reward_func": 0.28203125,
|
| 1233 |
+
"step": 510
|
| 1234 |
+
},
|
| 1235 |
+
{
|
| 1236 |
+
"completion_length": 167.35703125,
|
| 1237 |
+
"epoch": 2.068273092369478,
|
| 1238 |
+
"grad_norm": 0.4017987847328186,
|
| 1239 |
+
"kl": 0.0014677543425932527,
|
| 1240 |
+
"learning_rate": 5e-06,
|
| 1241 |
+
"loss": 0.0001,
|
| 1242 |
+
"reward": 0.36484375,
|
| 1243 |
+
"reward_std": 0.2759640544652939,
|
| 1244 |
+
"rewards/acc_reward_func": 0.36484375,
|
| 1245 |
+
"step": 515
|
| 1246 |
+
},
|
| 1247 |
+
{
|
| 1248 |
+
"completion_length": 171.8515625,
|
| 1249 |
+
"epoch": 2.0883534136546187,
|
| 1250 |
+
"grad_norm": 0.3457529842853546,
|
| 1251 |
+
"kl": 0.0014000870054587723,
|
| 1252 |
+
"learning_rate": 5e-06,
|
| 1253 |
+
"loss": 0.0001,
|
| 1254 |
+
"reward": 0.31875,
|
| 1255 |
+
"reward_std": 0.2790384829044342,
|
| 1256 |
+
"rewards/acc_reward_func": 0.31875,
|
| 1257 |
+
"step": 520
|
| 1258 |
+
},
|
| 1259 |
+
{
|
| 1260 |
+
"completion_length": 164.70078125,
|
| 1261 |
+
"epoch": 2.108433734939759,
|
| 1262 |
+
"grad_norm": 0.21619907021522522,
|
| 1263 |
+
"kl": 0.0014295668806880713,
|
| 1264 |
+
"learning_rate": 5e-06,
|
| 1265 |
+
"loss": 0.0001,
|
| 1266 |
+
"reward": 0.3140625,
|
| 1267 |
+
"reward_std": 0.2750910699367523,
|
| 1268 |
+
"rewards/acc_reward_func": 0.3140625,
|
| 1269 |
+
"step": 525
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"completion_length": 169.97265625,
|
| 1273 |
+
"epoch": 2.1285140562248994,
|
| 1274 |
+
"grad_norm": 0.31079721450805664,
|
| 1275 |
+
"kl": 0.0012559856520965695,
|
| 1276 |
+
"learning_rate": 5e-06,
|
| 1277 |
+
"loss": 0.0001,
|
| 1278 |
+
"reward": 0.33125,
|
| 1279 |
+
"reward_std": 0.24622083306312562,
|
| 1280 |
+
"rewards/acc_reward_func": 0.33125,
|
| 1281 |
+
"step": 530
|
| 1282 |
+
},
|
| 1283 |
+
{
|
| 1284 |
+
"completion_length": 170.13984375,
|
| 1285 |
+
"epoch": 2.1485943775100402,
|
| 1286 |
+
"grad_norm": 0.27532029151916504,
|
| 1287 |
+
"kl": 0.0015559423482045531,
|
| 1288 |
+
"learning_rate": 5e-06,
|
| 1289 |
+
"loss": 0.0002,
|
| 1290 |
+
"reward": 0.38359375,
|
| 1291 |
+
"reward_std": 0.28083202838897703,
|
| 1292 |
+
"rewards/acc_reward_func": 0.38359375,
|
| 1293 |
+
"step": 535
|
| 1294 |
+
},
|
| 1295 |
+
{
|
| 1296 |
+
"completion_length": 177.421875,
|
| 1297 |
+
"epoch": 2.1686746987951806,
|
| 1298 |
+
"grad_norm": 0.31572937965393066,
|
| 1299 |
+
"kl": 0.0013410489307716488,
|
| 1300 |
+
"learning_rate": 5e-06,
|
| 1301 |
+
"loss": 0.0001,
|
| 1302 |
+
"reward": 0.265625,
|
| 1303 |
+
"reward_std": 0.2276999294757843,
|
| 1304 |
+
"rewards/acc_reward_func": 0.265625,
|
| 1305 |
+
"step": 540
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"completion_length": 167.60234375,
|
| 1309 |
+
"epoch": 2.1887550200803214,
|
| 1310 |
+
"grad_norm": 0.29043304920196533,
|
| 1311 |
+
"kl": 0.0014236285351216793,
|
| 1312 |
+
"learning_rate": 5e-06,
|
| 1313 |
+
"loss": 0.0001,
|
| 1314 |
+
"reward": 0.3203125,
|
| 1315 |
+
"reward_std": 0.23548413515090943,
|
| 1316 |
+
"rewards/acc_reward_func": 0.3203125,
|
| 1317 |
+
"step": 545
|
| 1318 |
+
},
|
| 1319 |
+
{
|
| 1320 |
+
"completion_length": 166.23671875,
|
| 1321 |
+
"epoch": 2.208835341365462,
|
| 1322 |
+
"grad_norm": 0.6554353833198547,
|
| 1323 |
+
"kl": 0.0018002047901973129,
|
| 1324 |
+
"learning_rate": 5e-06,
|
| 1325 |
+
"loss": 0.0002,
|
| 1326 |
+
"reward": 0.3734375,
|
| 1327 |
+
"reward_std": 0.2930103540420532,
|
| 1328 |
+
"rewards/acc_reward_func": 0.3734375,
|
| 1329 |
+
"step": 550
|
| 1330 |
+
},
|
| 1331 |
+
{
|
| 1332 |
+
"completion_length": 159.91953125,
|
| 1333 |
+
"epoch": 2.2289156626506026,
|
| 1334 |
+
"grad_norm": 0.5253407955169678,
|
| 1335 |
+
"kl": 0.0016718338709324598,
|
| 1336 |
+
"learning_rate": 5e-06,
|
| 1337 |
+
"loss": 0.0002,
|
| 1338 |
+
"reward": 0.3703125,
|
| 1339 |
+
"reward_std": 0.2658358722925186,
|
| 1340 |
+
"rewards/acc_reward_func": 0.3703125,
|
| 1341 |
+
"step": 555
|
| 1342 |
+
},
|
| 1343 |
+
{
|
| 1344 |
+
"completion_length": 172.54296875,
|
| 1345 |
+
"epoch": 2.248995983935743,
|
| 1346 |
+
"grad_norm": 0.28389930725097656,
|
| 1347 |
+
"kl": 0.0020956686232239006,
|
| 1348 |
+
"learning_rate": 5e-06,
|
| 1349 |
+
"loss": 0.0002,
|
| 1350 |
+
"reward": 0.3203125,
|
| 1351 |
+
"reward_std": 0.2805899143218994,
|
| 1352 |
+
"rewards/acc_reward_func": 0.3203125,
|
| 1353 |
+
"step": 560
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"completion_length": 166.2515625,
|
| 1357 |
+
"epoch": 2.2690763052208833,
|
| 1358 |
+
"grad_norm": 0.27247288823127747,
|
| 1359 |
+
"kl": 0.0015082385856658221,
|
| 1360 |
+
"learning_rate": 5e-06,
|
| 1361 |
+
"loss": 0.0002,
|
| 1362 |
+
"reward": 0.3296875,
|
| 1363 |
+
"reward_std": 0.2336159199476242,
|
| 1364 |
+
"rewards/acc_reward_func": 0.3296875,
|
| 1365 |
+
"step": 565
|
| 1366 |
+
},
|
| 1367 |
+
{
|
| 1368 |
+
"completion_length": 157.78359375,
|
| 1369 |
+
"epoch": 2.289156626506024,
|
| 1370 |
+
"grad_norm": 0.31251490116119385,
|
| 1371 |
+
"kl": 0.001438130042515695,
|
| 1372 |
+
"learning_rate": 5e-06,
|
| 1373 |
+
"loss": 0.0001,
|
| 1374 |
+
"reward": 0.3796875,
|
| 1375 |
+
"reward_std": 0.277830982208252,
|
| 1376 |
+
"rewards/acc_reward_func": 0.3796875,
|
| 1377 |
+
"step": 570
|
| 1378 |
+
},
|
| 1379 |
+
{
|
| 1380 |
+
"completion_length": 172.43984375,
|
| 1381 |
+
"epoch": 2.3092369477911645,
|
| 1382 |
+
"grad_norm": 0.37210726737976074,
|
| 1383 |
+
"kl": 0.001831929781474173,
|
| 1384 |
+
"learning_rate": 5e-06,
|
| 1385 |
+
"loss": 0.0002,
|
| 1386 |
+
"reward": 0.296875,
|
| 1387 |
+
"reward_std": 0.2679367482662201,
|
| 1388 |
+
"rewards/acc_reward_func": 0.296875,
|
| 1389 |
+
"step": 575
|
| 1390 |
+
},
|
| 1391 |
+
{
|
| 1392 |
+
"completion_length": 162.16015625,
|
| 1393 |
+
"epoch": 2.3293172690763053,
|
| 1394 |
+
"grad_norm": 0.5997582077980042,
|
| 1395 |
+
"kl": 0.0017479128437116742,
|
| 1396 |
+
"learning_rate": 5e-06,
|
| 1397 |
+
"loss": 0.0002,
|
| 1398 |
+
"reward": 0.3875,
|
| 1399 |
+
"reward_std": 0.28567528128623965,
|
| 1400 |
+
"rewards/acc_reward_func": 0.3875,
|
| 1401 |
+
"step": 580
|
| 1402 |
+
},
|
| 1403 |
+
{
|
| 1404 |
+
"completion_length": 167.84765625,
|
| 1405 |
+
"epoch": 2.3493975903614457,
|
| 1406 |
+
"grad_norm": 0.23533369600772858,
|
| 1407 |
+
"kl": 0.001749329548329115,
|
| 1408 |
+
"learning_rate": 5e-06,
|
| 1409 |
+
"loss": 0.0002,
|
| 1410 |
+
"reward": 0.31875,
|
| 1411 |
+
"reward_std": 0.2516347885131836,
|
| 1412 |
+
"rewards/acc_reward_func": 0.31875,
|
| 1413 |
+
"step": 585
|
| 1414 |
+
},
|
| 1415 |
+
{
|
| 1416 |
+
"completion_length": 165.9859375,
|
| 1417 |
+
"epoch": 2.3694779116465865,
|
| 1418 |
+
"grad_norm": 0.3740471303462982,
|
| 1419 |
+
"kl": 0.001700690435245633,
|
| 1420 |
+
"learning_rate": 5e-06,
|
| 1421 |
+
"loss": 0.0002,
|
| 1422 |
+
"reward": 0.31484375,
|
| 1423 |
+
"reward_std": 0.24785314798355101,
|
| 1424 |
+
"rewards/acc_reward_func": 0.31484375,
|
| 1425 |
+
"step": 590
|
| 1426 |
+
},
|
| 1427 |
+
{
|
| 1428 |
+
"completion_length": 160.9109375,
|
| 1429 |
+
"epoch": 2.389558232931727,
|
| 1430 |
+
"grad_norm": 0.8214355111122131,
|
| 1431 |
+
"kl": 0.0020116518251597883,
|
| 1432 |
+
"learning_rate": 5e-06,
|
| 1433 |
+
"loss": 0.0002,
|
| 1434 |
+
"reward": 0.37265625,
|
| 1435 |
+
"reward_std": 0.2642554700374603,
|
| 1436 |
+
"rewards/acc_reward_func": 0.37265625,
|
| 1437 |
+
"step": 595
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"completion_length": 177.66796875,
|
| 1441 |
+
"epoch": 2.4096385542168672,
|
| 1442 |
+
"grad_norm": 0.48972687125205994,
|
| 1443 |
+
"kl": 0.0014791298424825072,
|
| 1444 |
+
"learning_rate": 5e-06,
|
| 1445 |
+
"loss": 0.0001,
|
| 1446 |
+
"reward": 0.2984375,
|
| 1447 |
+
"reward_std": 0.2614441394805908,
|
| 1448 |
+
"rewards/acc_reward_func": 0.2984375,
|
| 1449 |
+
"step": 600
|
| 1450 |
+
},
|
| 1451 |
+
{
|
| 1452 |
+
"completion_length": 167.84765625,
|
| 1453 |
+
"epoch": 2.429718875502008,
|
| 1454 |
+
"grad_norm": 0.29758477210998535,
|
| 1455 |
+
"kl": 0.0022905914578586818,
|
| 1456 |
+
"learning_rate": 5e-06,
|
| 1457 |
+
"loss": 0.0002,
|
| 1458 |
+
"reward": 0.321875,
|
| 1459 |
+
"reward_std": 0.2514772891998291,
|
| 1460 |
+
"rewards/acc_reward_func": 0.321875,
|
| 1461 |
+
"step": 605
|
| 1462 |
+
},
|
| 1463 |
+
{
|
| 1464 |
+
"completion_length": 160.82265625,
|
| 1465 |
+
"epoch": 2.4497991967871484,
|
| 1466 |
+
"grad_norm": 0.3575228750705719,
|
| 1467 |
+
"kl": 0.0013975306414067746,
|
| 1468 |
+
"learning_rate": 5e-06,
|
| 1469 |
+
"loss": 0.0001,
|
| 1470 |
+
"reward": 0.34765625,
|
| 1471 |
+
"reward_std": 0.28261533975601194,
|
| 1472 |
+
"rewards/acc_reward_func": 0.34765625,
|
| 1473 |
+
"step": 610
|
| 1474 |
+
},
|
| 1475 |
+
{
|
| 1476 |
+
"completion_length": 165.34296875,
|
| 1477 |
+
"epoch": 2.4698795180722892,
|
| 1478 |
+
"grad_norm": 0.21417830884456635,
|
| 1479 |
+
"kl": 0.0015371570363640786,
|
| 1480 |
+
"learning_rate": 5e-06,
|
| 1481 |
+
"loss": 0.0002,
|
| 1482 |
+
"reward": 0.33671875,
|
| 1483 |
+
"reward_std": 0.25463504195213316,
|
| 1484 |
+
"rewards/acc_reward_func": 0.33671875,
|
| 1485 |
+
"step": 615
|
| 1486 |
+
},
|
| 1487 |
+
{
|
| 1488 |
+
"completion_length": 173.79921875,
|
| 1489 |
+
"epoch": 2.4899598393574296,
|
| 1490 |
+
"grad_norm": 0.3487497866153717,
|
| 1491 |
+
"kl": 0.0016002030344679952,
|
| 1492 |
+
"learning_rate": 5e-06,
|
| 1493 |
+
"loss": 0.0002,
|
| 1494 |
+
"reward": 0.35078125,
|
| 1495 |
+
"reward_std": 0.2584144353866577,
|
| 1496 |
+
"rewards/acc_reward_func": 0.35078125,
|
| 1497 |
+
"step": 620
|
| 1498 |
+
},
|
| 1499 |
+
{
|
| 1500 |
+
"completion_length": 170.08125,
|
| 1501 |
+
"epoch": 2.5100401606425704,
|
| 1502 |
+
"grad_norm": 0.34159645438194275,
|
| 1503 |
+
"kl": 0.0013422498479485512,
|
| 1504 |
+
"learning_rate": 5e-06,
|
| 1505 |
+
"loss": 0.0001,
|
| 1506 |
+
"reward": 0.38203125,
|
| 1507 |
+
"reward_std": 0.2949507474899292,
|
| 1508 |
+
"rewards/acc_reward_func": 0.38203125,
|
| 1509 |
+
"step": 625
|
| 1510 |
+
},
|
| 1511 |
+
{
|
| 1512 |
+
"completion_length": 163.34765625,
|
| 1513 |
+
"epoch": 2.5301204819277108,
|
| 1514 |
+
"grad_norm": 0.44099095463752747,
|
| 1515 |
+
"kl": 0.0015945957973599433,
|
| 1516 |
+
"learning_rate": 5e-06,
|
| 1517 |
+
"loss": 0.0002,
|
| 1518 |
+
"reward": 0.34921875,
|
| 1519 |
+
"reward_std": 0.2757268697023392,
|
| 1520 |
+
"rewards/acc_reward_func": 0.34921875,
|
| 1521 |
+
"step": 630
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"completion_length": 170.05703125,
|
| 1525 |
+
"epoch": 2.550200803212851,
|
| 1526 |
+
"grad_norm": 0.4719444215297699,
|
| 1527 |
+
"kl": 0.0016027268255129456,
|
| 1528 |
+
"learning_rate": 5e-06,
|
| 1529 |
+
"loss": 0.0002,
|
| 1530 |
+
"reward": 0.32109375,
|
| 1531 |
+
"reward_std": 0.2545212864875793,
|
| 1532 |
+
"rewards/acc_reward_func": 0.32109375,
|
| 1533 |
+
"step": 635
|
| 1534 |
+
},
|
| 1535 |
+
{
|
| 1536 |
+
"completion_length": 167.87421875,
|
| 1537 |
+
"epoch": 2.570281124497992,
|
| 1538 |
+
"grad_norm": 0.34449702501296997,
|
| 1539 |
+
"kl": 0.001873347139917314,
|
| 1540 |
+
"learning_rate": 5e-06,
|
| 1541 |
+
"loss": 0.0002,
|
| 1542 |
+
"reward": 0.30390625,
|
| 1543 |
+
"reward_std": 0.2682854264974594,
|
| 1544 |
+
"rewards/acc_reward_func": 0.30390625,
|
| 1545 |
+
"step": 640
|
| 1546 |
+
},
|
| 1547 |
+
{
|
| 1548 |
+
"completion_length": 158.384375,
|
| 1549 |
+
"epoch": 2.5903614457831328,
|
| 1550 |
+
"grad_norm": 0.35067522525787354,
|
| 1551 |
+
"kl": 0.0016631773207336665,
|
| 1552 |
+
"learning_rate": 5e-06,
|
| 1553 |
+
"loss": 0.0002,
|
| 1554 |
+
"reward": 0.4421875,
|
| 1555 |
+
"reward_std": 0.3065200746059418,
|
| 1556 |
+
"rewards/acc_reward_func": 0.4421875,
|
| 1557 |
+
"step": 645
|
| 1558 |
+
},
|
| 1559 |
+
{
|
| 1560 |
+
"completion_length": 171.3046875,
|
| 1561 |
+
"epoch": 2.610441767068273,
|
| 1562 |
+
"grad_norm": 0.406903475522995,
|
| 1563 |
+
"kl": 0.001370012597180903,
|
| 1564 |
+
"learning_rate": 5e-06,
|
| 1565 |
+
"loss": 0.0001,
|
| 1566 |
+
"reward": 0.32265625,
|
| 1567 |
+
"reward_std": 0.2770428955554962,
|
| 1568 |
+
"rewards/acc_reward_func": 0.32265625,
|
| 1569 |
+
"step": 650
|
| 1570 |
+
},
|
| 1571 |
+
{
|
| 1572 |
+
"completion_length": 170.6921875,
|
| 1573 |
+
"epoch": 2.6305220883534135,
|
| 1574 |
+
"grad_norm": 0.3024086654186249,
|
| 1575 |
+
"kl": 0.0018637768691405654,
|
| 1576 |
+
"learning_rate": 5e-06,
|
| 1577 |
+
"loss": 0.0002,
|
| 1578 |
+
"reward": 0.33828125,
|
| 1579 |
+
"reward_std": 0.2950285911560059,
|
| 1580 |
+
"rewards/acc_reward_func": 0.33828125,
|
| 1581 |
+
"step": 655
|
| 1582 |
+
},
|
| 1583 |
+
{
|
| 1584 |
+
"completion_length": 161.92890625,
|
| 1585 |
+
"epoch": 2.6506024096385543,
|
| 1586 |
+
"grad_norm": 0.25132983922958374,
|
| 1587 |
+
"kl": 0.0013798804953694343,
|
| 1588 |
+
"learning_rate": 5e-06,
|
| 1589 |
+
"loss": 0.0001,
|
| 1590 |
+
"reward": 0.33671875,
|
| 1591 |
+
"reward_std": 0.25781014263629914,
|
| 1592 |
+
"rewards/acc_reward_func": 0.33671875,
|
| 1593 |
+
"step": 660
|
| 1594 |
+
},
|
| 1595 |
+
{
|
| 1596 |
+
"completion_length": 165.38203125,
|
| 1597 |
+
"epoch": 2.6706827309236947,
|
| 1598 |
+
"grad_norm": 0.24434784054756165,
|
| 1599 |
+
"kl": 0.0015540448017418384,
|
| 1600 |
+
"learning_rate": 5e-06,
|
| 1601 |
+
"loss": 0.0002,
|
| 1602 |
+
"reward": 0.34609375,
|
| 1603 |
+
"reward_std": 0.2942123174667358,
|
| 1604 |
+
"rewards/acc_reward_func": 0.34609375,
|
| 1605 |
+
"step": 665
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"completion_length": 168.7765625,
|
| 1609 |
+
"epoch": 2.6907630522088355,
|
| 1610 |
+
"grad_norm": 0.28431424498558044,
|
| 1611 |
+
"kl": 0.0027118735713884236,
|
| 1612 |
+
"learning_rate": 5e-06,
|
| 1613 |
+
"loss": 0.0003,
|
| 1614 |
+
"reward": 0.32734375,
|
| 1615 |
+
"reward_std": 0.26073101758956907,
|
| 1616 |
+
"rewards/acc_reward_func": 0.32734375,
|
| 1617 |
+
"step": 670
|
| 1618 |
+
},
|
| 1619 |
+
{
|
| 1620 |
+
"completion_length": 166.51484375,
|
| 1621 |
+
"epoch": 2.710843373493976,
|
| 1622 |
+
"grad_norm": 0.2701532542705536,
|
| 1623 |
+
"kl": 0.0017335619311779737,
|
| 1624 |
+
"learning_rate": 5e-06,
|
| 1625 |
+
"loss": 0.0002,
|
| 1626 |
+
"reward": 0.37890625,
|
| 1627 |
+
"reward_std": 0.29145594835281374,
|
| 1628 |
+
"rewards/acc_reward_func": 0.37890625,
|
| 1629 |
+
"step": 675
|
| 1630 |
+
},
|
| 1631 |
+
{
|
| 1632 |
+
"completion_length": 172.08359375,
|
| 1633 |
+
"epoch": 2.7309236947791167,
|
| 1634 |
+
"grad_norm": 0.25592124462127686,
|
| 1635 |
+
"kl": 0.001596282934769988,
|
| 1636 |
+
"learning_rate": 5e-06,
|
| 1637 |
+
"loss": 0.0002,
|
| 1638 |
+
"reward": 0.3484375,
|
| 1639 |
+
"reward_std": 0.3067091882228851,
|
| 1640 |
+
"rewards/acc_reward_func": 0.3484375,
|
| 1641 |
+
"step": 680
|
| 1642 |
+
},
|
| 1643 |
+
{
|
| 1644 |
+
"completion_length": 169.3625,
|
| 1645 |
+
"epoch": 2.751004016064257,
|
| 1646 |
+
"grad_norm": 0.369191437959671,
|
| 1647 |
+
"kl": 0.0018961878260597587,
|
| 1648 |
+
"learning_rate": 5e-06,
|
| 1649 |
+
"loss": 0.0002,
|
| 1650 |
+
"reward": 0.35703125,
|
| 1651 |
+
"reward_std": 0.26675377786159515,
|
| 1652 |
+
"rewards/acc_reward_func": 0.35703125,
|
| 1653 |
+
"step": 685
|
| 1654 |
+
},
|
| 1655 |
+
{
|
| 1656 |
+
"completion_length": 167.40546875,
|
| 1657 |
+
"epoch": 2.7710843373493974,
|
| 1658 |
+
"grad_norm": 0.3414689600467682,
|
| 1659 |
+
"kl": 0.001623287471011281,
|
| 1660 |
+
"learning_rate": 5e-06,
|
| 1661 |
+
"loss": 0.0002,
|
| 1662 |
+
"reward": 0.34375,
|
| 1663 |
+
"reward_std": 0.24926708936691283,
|
| 1664 |
+
"rewards/acc_reward_func": 0.34375,
|
| 1665 |
+
"step": 690
|
| 1666 |
+
},
|
| 1667 |
+
{
|
| 1668 |
+
"completion_length": 164.5421875,
|
| 1669 |
+
"epoch": 2.791164658634538,
|
| 1670 |
+
"grad_norm": 0.2602121829986572,
|
| 1671 |
+
"kl": 0.0020977890817448497,
|
| 1672 |
+
"learning_rate": 5e-06,
|
| 1673 |
+
"loss": 0.0002,
|
| 1674 |
+
"reward": 0.3359375,
|
| 1675 |
+
"reward_std": 0.2536847472190857,
|
| 1676 |
+
"rewards/acc_reward_func": 0.3359375,
|
| 1677 |
+
"step": 695
|
| 1678 |
+
},
|
| 1679 |
+
{
|
| 1680 |
+
"completion_length": 153.6890625,
|
| 1681 |
+
"epoch": 2.8112449799196786,
|
| 1682 |
+
"grad_norm": 0.28262779116630554,
|
| 1683 |
+
"kl": 0.0015694845002144574,
|
| 1684 |
+
"learning_rate": 5e-06,
|
| 1685 |
+
"loss": 0.0002,
|
| 1686 |
+
"reward": 0.38046875,
|
| 1687 |
+
"reward_std": 0.273573100566864,
|
| 1688 |
+
"rewards/acc_reward_func": 0.38046875,
|
| 1689 |
+
"step": 700
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"completion_length": 166.97421875,
|
| 1693 |
+
"epoch": 2.8313253012048194,
|
| 1694 |
+
"grad_norm": 0.2816270589828491,
|
| 1695 |
+
"kl": 0.001585571584291756,
|
| 1696 |
+
"learning_rate": 5e-06,
|
| 1697 |
+
"loss": 0.0002,
|
| 1698 |
+
"reward": 0.36640625,
|
| 1699 |
+
"reward_std": 0.30526658296585085,
|
| 1700 |
+
"rewards/acc_reward_func": 0.36640625,
|
| 1701 |
+
"step": 705
|
| 1702 |
+
},
|
| 1703 |
+
{
|
| 1704 |
+
"completion_length": 163.82421875,
|
| 1705 |
+
"epoch": 2.8514056224899598,
|
| 1706 |
+
"grad_norm": 0.35617348551750183,
|
| 1707 |
+
"kl": 0.0016123745823279022,
|
| 1708 |
+
"learning_rate": 5e-06,
|
| 1709 |
+
"loss": 0.0002,
|
| 1710 |
+
"reward": 0.34375,
|
| 1711 |
+
"reward_std": 0.26365512013435366,
|
| 1712 |
+
"rewards/acc_reward_func": 0.34375,
|
| 1713 |
+
"step": 710
|
| 1714 |
+
},
|
| 1715 |
+
{
|
| 1716 |
+
"completion_length": 168.128125,
|
| 1717 |
+
"epoch": 2.8714859437751006,
|
| 1718 |
+
"grad_norm": 0.20444567501544952,
|
| 1719 |
+
"kl": 0.001708123623393476,
|
| 1720 |
+
"learning_rate": 5e-06,
|
| 1721 |
+
"loss": 0.0002,
|
| 1722 |
+
"reward": 0.378125,
|
| 1723 |
+
"reward_std": 0.27598778903484344,
|
| 1724 |
+
"rewards/acc_reward_func": 0.378125,
|
| 1725 |
+
"step": 715
|
| 1726 |
+
},
|
| 1727 |
+
{
|
| 1728 |
+
"completion_length": 158.5953125,
|
| 1729 |
+
"epoch": 2.891566265060241,
|
| 1730 |
+
"grad_norm": 0.23954364657402039,
|
| 1731 |
+
"kl": 0.001798305264674127,
|
| 1732 |
+
"learning_rate": 5e-06,
|
| 1733 |
+
"loss": 0.0002,
|
| 1734 |
+
"reward": 0.36875,
|
| 1735 |
+
"reward_std": 0.28262283504009245,
|
| 1736 |
+
"rewards/acc_reward_func": 0.36875,
|
| 1737 |
+
"step": 720
|
| 1738 |
+
},
|
| 1739 |
+
{
|
| 1740 |
+
"completion_length": 168.59453125,
|
| 1741 |
+
"epoch": 2.9116465863453813,
|
| 1742 |
+
"grad_norm": 0.5855829119682312,
|
| 1743 |
+
"kl": 0.0019487401703372597,
|
| 1744 |
+
"learning_rate": 5e-06,
|
| 1745 |
+
"loss": 0.0002,
|
| 1746 |
+
"reward": 0.33046875,
|
| 1747 |
+
"reward_std": 0.26820210814476014,
|
| 1748 |
+
"rewards/acc_reward_func": 0.33046875,
|
| 1749 |
+
"step": 725
|
| 1750 |
+
},
|
| 1751 |
+
{
|
| 1752 |
+
"completion_length": 173.31171875,
|
| 1753 |
+
"epoch": 2.931726907630522,
|
| 1754 |
+
"grad_norm": 0.4006167948246002,
|
| 1755 |
+
"kl": 0.0018504543462768198,
|
| 1756 |
+
"learning_rate": 5e-06,
|
| 1757 |
+
"loss": 0.0002,
|
| 1758 |
+
"reward": 0.3234375,
|
| 1759 |
+
"reward_std": 0.2933495879173279,
|
| 1760 |
+
"rewards/acc_reward_func": 0.3234375,
|
| 1761 |
+
"step": 730
|
| 1762 |
+
},
|
| 1763 |
+
{
|
| 1764 |
+
"completion_length": 168.96015625,
|
| 1765 |
+
"epoch": 2.9518072289156625,
|
| 1766 |
+
"grad_norm": 0.4413718581199646,
|
| 1767 |
+
"kl": 0.002411281201057136,
|
| 1768 |
+
"learning_rate": 5e-06,
|
| 1769 |
+
"loss": 0.0002,
|
| 1770 |
+
"reward": 0.34765625,
|
| 1771 |
+
"reward_std": 0.2572381556034088,
|
| 1772 |
+
"rewards/acc_reward_func": 0.34765625,
|
| 1773 |
+
"step": 735
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"completion_length": 172.9421875,
|
| 1777 |
+
"epoch": 2.9718875502008033,
|
| 1778 |
+
"grad_norm": 0.3350299596786499,
|
| 1779 |
+
"kl": 0.002306809718720615,
|
| 1780 |
+
"learning_rate": 5e-06,
|
| 1781 |
+
"loss": 0.0002,
|
| 1782 |
+
"reward": 0.32109375,
|
| 1783 |
+
"reward_std": 0.27035961151123045,
|
| 1784 |
+
"rewards/acc_reward_func": 0.32109375,
|
| 1785 |
+
"step": 740
|
| 1786 |
+
},
|
| 1787 |
+
{
|
| 1788 |
+
"completion_length": 169.828125,
|
| 1789 |
+
"epoch": 2.9919678714859437,
|
| 1790 |
+
"grad_norm": 0.4135463535785675,
|
| 1791 |
+
"kl": 0.002015371876768768,
|
| 1792 |
+
"learning_rate": 5e-06,
|
| 1793 |
+
"loss": 0.0002,
|
| 1794 |
+
"reward": 0.3859375,
|
| 1795 |
+
"reward_std": 0.3136798143386841,
|
| 1796 |
+
"rewards/acc_reward_func": 0.3859375,
|
| 1797 |
+
"step": 745
|
| 1798 |
+
},
|
| 1799 |
+
{
|
| 1800 |
+
"completion_length": 185.21177673339844,
|
| 1801 |
+
"epoch": 3.0,
|
| 1802 |
+
"kl": 0.0030494448728859425,
|
| 1803 |
+
"reward": 0.267578125,
|
| 1804 |
+
"reward_std": 0.2742668390274048,
|
| 1805 |
+
"rewards/acc_reward_func": 0.267578125,
|
| 1806 |
+
"step": 747,
|
| 1807 |
+
"total_flos": 0.0,
|
| 1808 |
+
"train_loss": 0.00012346780326337724,
|
| 1809 |
+
"train_runtime": 38700.7369,
|
| 1810 |
+
"train_samples_per_second": 0.616,
|
| 1811 |
+
"train_steps_per_second": 0.019
|
| 1812 |
+
}
|
| 1813 |
+
],
|
| 1814 |
+
"logging_steps": 5,
|
| 1815 |
+
"max_steps": 747,
|
| 1816 |
+
"num_input_tokens_seen": 0,
|
| 1817 |
+
"num_train_epochs": 3,
|
| 1818 |
+
"save_steps": 500,
|
| 1819 |
+
"stateful_callbacks": {
|
| 1820 |
+
"TrainerControl": {
|
| 1821 |
+
"args": {
|
| 1822 |
+
"should_epoch_stop": false,
|
| 1823 |
+
"should_evaluate": false,
|
| 1824 |
+
"should_log": false,
|
| 1825 |
+
"should_save": true,
|
| 1826 |
+
"should_training_stop": true
|
| 1827 |
+
},
|
| 1828 |
+
"attributes": {}
|
| 1829 |
+
}
|
| 1830 |
+
},
|
| 1831 |
+
"total_flos": 0.0,
|
| 1832 |
+
"train_batch_size": 64,
|
| 1833 |
+
"trial_name": null,
|
| 1834 |
+
"trial_params": null
|
| 1835 |
+
}
|