34 lines
922 B
Python
34 lines
922 B
Python
|
import warnings
|
||
|
|
|
||
|
|
import gradio as gr
|
||
|
|
import openai
|
||
|
|
|
||
|
|
warnings.filterwarnings("ignore")
|
||
|
|
|
||
|
|
openai.api_key = 'null'
|
||
|
|
openai.api_base = 'http://localhost:5000/api/openai/v1'
|
||
|
|
|
||
|
|
|
||
|
|
def stream_response(prompt, history):
|
||
|
|
messages = []
|
||
|
|
for x in history:
|
||
|
|
messages.append({'role': 'user', 'content': x[0]})
|
||
|
|
messages.append({'role': 'assistant', 'content': x[1]})
|
||
|
|
messages.append({'role': 'user', 'content': prompt})
|
||
|
|
|
||
|
|
response = openai.ChatCompletion.create(
|
||
|
|
model='0',
|
||
|
|
messages=messages,
|
||
|
|
temperature=0,
|
||
|
|
max_tokens=300,
|
||
|
|
stream=True
|
||
|
|
)
|
||
|
|
|
||
|
|
message = ''
|
||
|
|
for chunk in response:
|
||
|
|
message += chunk['choices'][0]['delta']['content']
|
||
|
|
yield message
|
||
|
|
|
||
|
|
|
||
|
|
gr.ChatInterface(stream_response, examples=["hello", "hola", "merhaba"], title="Chatbot Demo", analytics_enabled=False, cache_examples=False, css='#component-0{height:100%!important}').queue().launch()
|