app.py

from transformers import AutoModel, AutoTokenizer
tokenizer = AutoTokenizer.from_pretrained("THUDM/chatglm2-6b", trust_remote_code=True)
model = AutoModel.from_pretrained("THUDM/chatglm2-6b", trust_remote_code=True).half().cuda()
model = model.eval()

def beginchat(input, history=None):
    if history is None:
        history = []
    response, history = model.chat(tokenizer, input, history)
    return history, history


with gr.Blocks() as chatglm2bot:
    gr.Markdown('''## ChatGLM2-6B - chatbot demo''')
    state = gr.State([])
    chatbot = gr.Chatbot([], elem_id="chatbot").style(height=400)
    with gr.Row():
        with gr.Column(scale=4):
            txt = gr.Textbox(show_label=False, placeholder="Enter text and press enter").style(container=False)
        with gr.Column(scale=1):
            button = gr.Button("Generate")
    txt.submit(beginchat, [txt, state], [chatbot, state])
    button.click(beginchat, [txt, state], [chatbot, state])
chatglm2bot.queue().launch()