Delete chat-qwen.py

chat-qwen.py

@@ -1,33 +0,0 @@
-from transformers import AutoModelForCausalLM, AutoTokenizer
-
-def get_response(prompt: str):
-    model = AutoModelForCausalLM.from_pretrained(
-        "Qwen/Qwen2-7B-Instruct-GPTQ-Int4",
-        torch_dtype="auto",
-        device_map="auto",
-    )
-
-    tokenizer = AutoTokenizer.from_pretrained("Qwen/Qwen2-7B-Instruct-GPTQ-Int4")
-
-    prompt = "Give me a short introduction to large language model."
-    messages = [
-        {"role": "system", "content": "You are a helpful assistant."},
-        {"role": "user", "content": prompt},
-    ]
-    text = tokenizer.apply_chat_template(
-        messages,
-        tokenize=False,
-        add_generation_prompt=True,
-    )
-    model_inputs = tokenizer([text], return_tensors="pt").to(model.device)
-
-    generated_ids = model.generate(
-        **model_inputs,
-        max_new_tokens=512,
-    )
-    generated_ids = [
-        output_ids[len(input_ids):] for input_ids, output_ids in zip(model_inputs.input_ids, generated_ids)
-    ]
-
-    response = tokenizer.batch_decode(generated_ids, skip_special_tokens=True)[0]
-    return response