Update inference_fine_tune.py

inference_fine_tune.py

@@ -22,7 +22,7 @@ def generate_text(
 
     # Print the initial prompt
 
-    while decoder_input.shape[1] < 2000 :
+    while decoder_input.shape[1] < 300 :
         # Apply causal mask based on current decoder_input length
         # decoder_mask = (decoder_input != pad_idx).unsqueeze(0).int() & causal_mask(decoder_input.size(1)).type_as(mask).to(device)