Update app.py

app.py

@@ -30,14 +30,14 @@ tokenizer = AutoTokenizer.from_pretrained(model_name)
 model = AutoModelForCausalLM.from_pretrained(
     model_name, 
     device_map=device,
-    torch_dtype=torch.float32 if device == "cuda" else torch.float32
+    torch_dtype=torch.float16 if device == "cuda" else torch.float32
 )
 #model = AutoModelForCausalLM.from_pretrained(model_name, device_map="auto", torch_dtype="auto")
 qa_pipeline = pipeline(
     "text-generation",
     model=model,
     tokenizer=tokenizer,
-    max_new_tokens=500,
+    max_new_tokens=1500,
     pad_token_id=tokenizer.eos_token_id
 )