Update app.py

app.py

@@ -20,10 +20,10 @@ rm_tokenizer = AutoTokenizer.from_pretrained('OpenAssistant/reward-model-deberta
 rm_model = AutoModelForSequenceClassification.from_pretrained('OpenAssistant/reward-model-deberta-v3-large-v2', torch_dtype=torch.bfloat16)
 
 @spaces.GPU
-def generate_text(usertitle, content, max_length, temperature, N=3):
+def generate_text(usertitle, content, temperature, max_length, N=3):
     input_text = f"title: {usertitle}\ncontent: {content}"
     inputs = tokenizer(input_text, return_tensors='pt').to('cuda')
-    generated_sequences = model.generate(inputs['input_ids'], max_new_tokens=max_length, temperature=temperature, num_return_sequences=N, do_sample=True)
+    generated_sequences = model.generate(inputs['input_ids'], attention_mask=attention_mask, temperature=temperature, max_length=max_length, pad_token_id=tokenizer.eos_token_id, num_return_sequences=N, do_sample=True)
     decoded_sequences = [tokenizer.decode(g, skip_special_tokens=True) for g in generated_sequences]
 
     def score(sequence):