Update README.md

README.md

@@ -17,7 +17,9 @@ In addition to cross entropy and cosine teacher-student losses, DistilProtBert w
 DistilProtBert was pretrained on millions of proteins sequences. 
 
 Few important differences between DistilProtBert model and the original ProtBert version are:
-1. Size of the model: 230M parameters (ProtBert has 420M parameters)
+1. Size of the model: 
+  - 230M parameters (ProtBert has 420M parameters)
+  - 15 hidden layers (ProtBert has 30 hidden layers)
 2. Size of the pretraining dataset: ~43M proteins (ProtBert was pretrained on 216M proteins)
 3. Hardware used for pretraining: five v100 32GB Nvidia GPUs (ProtBert was pretrained on 512 16GB TPUs)