Remove infra details from model card
Browse files
README.md
CHANGED
|
@@ -40,7 +40,7 @@ RL-trained Qwen3-8B on SWEsmith tasks (32k context, no rope scaling, 35 steps).
|
|
| 40 |
| **Policy nodes** | 2 (8 GPUs, FSDP2) |
|
| 41 |
| **Inference engines** | 20 (TP=1) |
|
| 42 |
| **Training steps** | 35 |
|
| 43 |
-
| **Framework** | BenSkyRL + Harbor
|
| 44 |
|
| 45 |
## Usage
|
| 46 |
|
|
|
|
| 40 |
| **Policy nodes** | 2 (8 GPUs, FSDP2) |
|
| 41 |
| **Inference engines** | 20 (TP=1) |
|
| 42 |
| **Training steps** | 35 |
|
| 43 |
+
| **Framework** | BenSkyRL + Harbor |
|
| 44 |
|
| 45 |
## Usage
|
| 46 |
|