Update model.py

llama/model.py

@@ -448,7 +448,8 @@ class Transformer(nn.Module):
         )
 
         self.freqs_cis = precompute_freqs_cis(
-            # self.params.max_seq_len is multiplied by 2 because the token limit is 4096, so you can't make it bigger than that, and this is why it's set the the maximun (4096).
+            # Note that self.params.max_seq_len is multiplied by 2 because the token limit for the Llama 2 generation of models is 4096. 
+            #Adding this multiplier instead of using 4096 directly allows for dynamism of token lengths while training or fine-tuning.
             self.params.dim // self.params.n_heads, self.params.max_seq_len * 2
         )