Update README.md
Browse files
README.md
CHANGED
|
@@ -31,12 +31,6 @@ JustGRPO is a minimalist RL approach for diffusion language models. Instead of c
|
|
| 31 |
|:---:|:---:|:---:|:---:|
|
| 32 |
| **Accuracy (%)** | 39.0 | 45.1 | 45.2 |
|
| 33 |
|
| 34 |
-
### GSM8K
|
| 35 |
-
|
| 36 |
-
| Sequence Length | 128 | 256 | 512 |
|
| 37 |
-
|:---:|:---:|:---:|:---:|
|
| 38 |
-
| **Accuracy (%)** | 83.8 | 89.1 | 89.8 |
|
| 39 |
-
|
| 40 |
## Usage
|
| 41 |
|
| 42 |
For generation and evaluation, please refer to our [GitHub repository](https://github.com/LeapLabTHU/JustGRPO).
|
|
|
|
| 31 |
|:---:|:---:|:---:|:---:|
|
| 32 |
| **Accuracy (%)** | 39.0 | 45.1 | 45.2 |
|
| 33 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 34 |
## Usage
|
| 35 |
|
| 36 |
For generation and evaluation, please refer to our [GitHub repository](https://github.com/LeapLabTHU/JustGRPO).
|