Update README.md

README.md

@@ -2,13 +2,11 @@
 This reward function can be used for RLHF, including PPO, iterative SFT, iterative DPO.
 
 ## Training
-The base model is meta-llama/Meta-Llama-3-8B-Instruct.
+The base model is `meta-llama/Meta-Llama-3-8B-Instruct`.
 
 We use the training script at `https://github.com/WeiXiongUST/RLHF-Reward-Modeling`.
 
 
-We train the model for one epoch with a learning rate of 2e-6, batch size 512, cosine learning rate decay with a warmup ratio 0.03.
-
 ## Uses
 
 ```python
@@ -45,7 +43,7 @@ We train the model for one epoch with a learning rate of 2e-6, batch size 512, c
 ## Results
 
 
-This Reward model is the SOTA open-source RM (Apr 20, 2024).
+This Reward model is the SOTA open-source RM (Apr 20, 2024) on Reward-Bench.
 
 | Metric       | Score  |
 |--------------|--------|