Update README.md

README.md

@@ -12,11 +12,11 @@ pipeline_tag: text-generation
 
 ![wsd_scheduler](./wsd_scheduler.png)
 
-Doge uses `wsd_scheduler` as the training scheduler, which divides the learning rate into three stages: warmup, stable, and decay. It allows us to continue training from any checkpoint in the stable stage without causing loss rebound.
+Doge uses `wsd_scheduler` as the training scheduler, which divides the learning rate into three stages: `warmup`, `stable`, and `decay`. It allows us to continue training from any checkpoint in the `stable stage` without causing loss rebound.
 
 Here are the initial learning rates required to continue training at each checkpoint:
 
-- `Doge-20M`: 8e-3
-- `Doge-60M`: 6e-3
-- `Doge-160M`: 4e-3
-- `Doge-320M`: 2e-3
+- **Doge-20M**: 8e-3
+- **Doge-60M**: 6e-3
+- **Doge-160M**: 4e-3
+- **Doge-320M**: 2e-3