---
library_name: transformers
tags:
- ppo
base_model: lvwerra/gpt2-imdb
datasets:
- imdb
---

# GPT2-IMDB

## Results

Reward Model: [lvwerra/distilbert-imdb](https://huggingface.co/lvwerra/distilbert-imdb)


| Statistic | Rewards (Before) | Rewards (After) |
|-----------|------------------|-----------------|
| Mean      | 1.147794         | 2.635067        |
| Median    | 1.525790         | 2.777761        |