Update README.md

README.md

@@ -41,7 +41,7 @@ GPT-J 6B was pretrained on the [Pile](pile.eleuther.ai), a large scale curated d
 
 ## Training procedure
 
-Genji-python-6B is trained for 20k steps on around 655 million tokens. For LR we used a warmup steps of 2000 and annual decrease of learning rate from 5e-06 to 1e-06.
+Genji-python-6B is trained for 20k steps on around 655 million tokens with learning rate of 2e-06
 
 ## Intended Use
 
@@ -94,7 +94,7 @@ Generation:
 ```
 
 For example usage, you can see our colab notebook as well:
-[Notebook]()
+[Notebook](https://colab.research.google.com/drive/1PnWpx02IEUkY8jhLKd_NewUGEXahAska?usp=sharing)
 
 ## Eval results