Update app.py

app.py

@@ -20,7 +20,7 @@ def make_inference(resume):
   batch = tokenizer(f"Write an excecutive summary based off this resume.\n\n### Resume:\n{resume}", return_tensors='pt')
 
   with torch.cuda.amp.autocast():
-    output_tokens = model.generate(**batch, max_new_tokens=50)
+    output_tokens = model.generate(**batch, max_new_tokens=200)
 
   return tokenizer.decode(output_tokens[0], skip_special_tokens=True)