File size: 1,984 Bytes
bd5df33 0348877 bd5df33 c18814e e44990b 0348877 c18814e 0348877 bd5df33 0348877 e44990b 0348877 e44990b 0348877 e44990b 0348877 e44990b 0348877 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 |
import gradio as gr
from transformers import AutoModel, AutoTokenizer
# Load the model
model_name = "wop/kosmox-gguf"
model = AutoModel.from_pretrained(model_name)
# Assuming we need to load a corresponding tokenizer
tokenizer = AutoTokenizer.from_pretrained(model_name)
# Define the chat template
chat_template = "{{ bos_token }}{% for message in messages %}{% if message['from'] == 'human' %}{{' ' + message['value'] + ' '}}{% elif message['from'] == 'gpt' %}{{' ' + message['value'] + ' '}}{% else %}{{'<|' + message['from'] + '|> ' + message['value'] + ' '}}{% endif %}{% endfor %}{% if add_generation_prompt %}{{ ' ' }}{% endif %}"
def chat(messages):
# Prepare the input
context = chat_template.format(
bos_token=tokenizer.bos_token,
messages=messages,
add_generation_prompt=True
)
# Tokenize the input
inputs = tokenizer(context, return_tensors='pt')
# Generate response
outputs = model.generate(**inputs)
# Decode the response
response = tokenizer.decode(outputs[0], skip_special_tokens=True)
return response
# Define the Gradio interface
with gr.Blocks() as demo:
chatbot = gr.Chatbot()
with gr.Row():
with gr.Column():
user_input = gr.Textbox(
placeholder="Type your message here...",
label="Your message"
)
send_button = gr.Button("Send")
with gr.Column():
chat_output = gr.Textbox(
label="Chatbot response",
interactive=False
)
def respond(message, history):
history = history or []
history.append({"from": "human", "value": message})
response = chat(history)
history.append({"from": "gpt", "value": response})
return history, history[-1]['value']
send_button.click(respond, [user_input, chatbot], [chatbot, chat_output])
# Launch the Gradio interface
demo.launch()
|