dev

app.py

@@ -256,10 +256,11 @@ def predict(text):
     if torch.cuda.is_available():
         model = model.to('cuda')
         inputs = {key: value.to('cuda') for key, value in inputs.items()}
-
+    print('xxxxxxxxxxxxxxxxxxxxxxx')
+    print(inputs)
     # Generate outputs
     with torch.no_grad():  # Disable gradient calculation for inference
-        outputs = model.generate(inputs)
+        outputs = model.generate(inputs['input_ids'][0])
 
     ## Decode the generated output
     #predictions = tokenizer.decode(outputs[0], skip_special_tokens=True)