Update README.md

README.md

@@ -51,7 +51,7 @@ Here's what the train/eval loss looked like:
 
 <img src="https://cdn-uploads.huggingface.co/production/uploads/662c17b252e194d5d436c708/hUKuy7ht_qObuFNDTVEe9.png" width="800"/>
 
-I still don't know what makes training loss drop at the end of epoch 1, or why eval loss doesn't drop down to match (it continues to decrease, but slowly). I did say this was experimental, right?
+I still don't know what makes training loss drop at the end of epoch 1, or why eval loss doesn't drop down to match (it continues to decrease, but slowly). I did say this was experimental, right? If I want to throw more money at this I might try a 3 epoch run just to see what happens.
 
 It was trained with the following settings: