Update app.py

app.py

@@ -94,16 +94,9 @@ def evaluate(
         num_beams=num_beams,
         **kwargs,
     )
-    with torch.no_grad():
-        generation_output = model.generate(
-            input_ids=input_ids,
-            generation_config=generation_config,
-            return_dict_in_generate=True,
-            output_scores=True,
-            max_new_tokens=max_new_tokens,
-        )
-    s = generation_output.sequences[0]
-    output = tokenizer.decode(s)
+    with torch.cuda.amp.autocast():
+        output_tokens = model.generate(**inputs, generation_config=generation_config)
+    output = tokenizer.decode(output_tokens[0], skip_special_tokens=True)
     return output[len(input_ids):]