Spaces:
Running
Running
File size: 2,701 Bytes
26ca9d4 9423469 4ef74d7 2493f19 4ef74d7 f00ccef 4ef74d7 f00ccef 628c773 26ca9d4 ebac2a9 26ca9d4 f00ccef 291372b 26ca9d4 9423469 f00ccef 291372b f00ccef eb16940 9423469 73a2adf 9f21eff 77b7aca 9f21eff cb6cb3b 9f21eff ebac2a9 9f21eff cb6cb3b 77b7aca cb6cb3b 9f21eff cb6cb3b 9f21eff 628c773 9f21eff 73a2adf 628c773 9f21eff |
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 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 |
# app.py
from transformers import AutoTokenizer, AutoModelForCausalLM, TextIteratorStreamer
from threading import Thread
import gradio as gr
import re
import torch
from openai import OpenAI
client = OpenAI(
api_key="sk-420ab66020704eabbe37501ec39b7a2b",
base_url="https://bailingchat.alipay.com",
)
# define chat function
def chat(user_input, max_new_tokens=20480):
# chat history
messages_template = [
# {"role": "system", "content": "You are Ling, an assistant created by inclusionAI"},
{"role": "user", "content": user_input}
]
response = client.chat.completions.create(
model="Ling-lite-1.5-250604",
messages=messages_template,
max_tokens=11264,
temperature=0.01,
top_p=1,
)
resp_text = response.choices[0].message.content
print(resp_text)
yield resp_text
# Create a custom layout using Blocks
with gr.Blocks(css="""
#markdown-output {
height: 300px;
overflow-y: auto;
border: 1px solid #ddd;
padding: 10px;
}
""") as demo:
gr.Markdown(
"## Ling-lite-1.5 AI Assistant\n"
"Based on [inclusionAI/Ling-lite-1.5](https://huggingface.co/inclusionAI/Ling-lite-1.5) "
)
with gr.Row():
max_tokens_slider = gr.Slider(minimum=128, maximum=20480, step=160, label="Generated length")
# output_box = gr.Textbox(lines=10, label="Response")
output_box = gr.Markdown(label="Response", elem_id="markdown-output")
input_box = gr.Textbox(lines=8, label="Input you question")
examples = gr.Examples(
examples=[
["Introducing the basic concepts of large language models"],
["How to solve long context dependencies in math problems?"]
],
inputs=input_box
)
interface = gr.Interface(
fn=chat,
inputs=[input_box, max_tokens_slider],
outputs=output_box,
live=False # disable auto-triggering on input change
)
# launch Gradio Service
demo.queue()
demo.launch()
# Construct Gradio Interface
#interface = gr.Interface(
# fn=chat,
# inputs=[
# gr.Textbox(lines=8, label="输入你的问题"),
# gr.Slider(minimum=100, maximum=102400, step=50, label="生成长度")
# ],
# outputs=[
# gr.Textbox(lines=8, label="模型回复")
# ],
# title="Ling-lite-1.5 AI助手",
# description="基于 [inclusionAI/Ling-lite-1.5](https://huggingface.co/inclusionAI/Ling-lite-1.5) 的对话式文本生成演示。",
# examples=[
# ["介绍大型语言模型的基本概念"],
# ["如何解决数学问题中的长上下文依赖？"]
# ]
#)
# launch Gradion Service
#interface.launch()
|