first

app.py

@@ -6,7 +6,6 @@ import os
 model_path = "vinai/PhoGPT-7B5-Instruct"
 
 config = AutoConfig.from_pretrained(model_path, trust_remote_code=True, token=os.environ['HK_TOKEN'])
-config.init_device = "cuda"
 # config.attn_config['attn_impl'] = 'triton' # Enable if "triton" installed!
 
 model = AutoModelForCausalLM.from_pretrained(
@@ -23,8 +22,8 @@ def answer(input_prompt):
     input_ids = tokenizer(input_prompt, return_tensors="pt")
 
     outputs = model.generate(
-        inputs=input_ids["input_ids"].to("cuda"),
-        attention_mask=input_ids["attention_mask"].to("cuda"),
+        inputs=input_ids["input_ids"].to("cpu"),
+        attention_mask=input_ids["attention_mask"].to("cpu"),
         do_sample=True,
         temperature=1.0,
         top_k=50,