Update app.py

app.py

@@ -22,7 +22,7 @@ def make_inference(abstract):
     batch = tokenizer(input_str, return_tensors="pt")
 
     with torch.cuda.amp.autocast():
-        output_tokens = model.generate(**batch, max_new_tokens=200)
+        output_tokens = model.generate(**batch, max_new_tokens=50)
 
     return tokenizer.decode(output_tokens[0], skip_special_tokens=True)