app.py

from transformers import AutoModelForCausalLM, AutoTokenizer
import os

hf_token = os.environ.get("HF_TOKEN")

model = AutoModelForCausalLM.from_pretrained(
    "Qwen/CodeQwen1.5-7B-Chat",
    torch_dtype="auto",
    device_map="auto",
    token=hf_token
)
tokenizer = AutoTokenizer.from_pretrained("Qwen/CodeQwen1.5-7B-Chat", token=hf_token)

messages = [
    {"role": "system", "content": "You are a helpful assistant."},
]
import gradio as gr

def greet(prompt):
    messages.append({"role": "user", "content": prompt})
    text = tokenizer.apply_chat_template(
    messages,
    tokenize=False,
    add_generation_prompt=True
    )
    model_inputs = tokenizer([text], return_tensors="pt")

    generated_ids = model.generate(
        model_inputs.input_ids,
        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].text
    
    messages.append({"role": "bot", "content": response})

    return response

demo = gr.Interface(fn=greet, inputs="text", outputs="text")
demo.launch()