一键部署Huggingface模型工具Gradio - 文章 - 开发者社区

使用过Huggingface的开发者，应该见过如下的界面，方便用户在文本框直接输入query，然后使用模型来预测输出答案

picture.image

接下来，我们看一下实现步骤（使用的部署工具是python包gradio）：

Note：部署工具可以在python任意的IDE运行，比如pycharm、jupyter notebook以及Google Colab，下面以Google Colab为例进行展示GPT-2模型的部署步骤：

Step1 ：首先就是安装python包


          
!pip install gradio
          
!pip install transformers

Step2 ：为了便于理解，我们首先实现一个输入文本，输出文本的demo，代码如下：


          
# Let’s get started with a simple “Hello World”
          
import gradio as gr
          

          
def greet(name):
          
    return "Hello " + name
          

          
# We instantiate the Textbox class
          
demo = gr.Interface(fn=greet, inputs="text", outputs="text")
          

          
demo.launch()

运行上述脚本，会生成如下界面：

picture.image

用户在name文本框中输入文本“x”，会在output文本框中生成“Hello x”文本内容

picture.image

Step3 ：当然以上demo也是可以进行润色的


          
# Upgrade “Hello World” 
          
import gradio as gr
          

          
def greet(name):
          
    return "Hello " + name
          

          
# We instantiate the Textbox class
          
textbox_input = gr.Textbox(label="Type your name here:", placeholder="John Doe", lines=2)
          
textbox2_output = gr.Textbox(label="View output result here:", placeholder="Arron", lines=2)
          
demo = gr.Interface(fn=greet, inputs=textbox_input, outputs=textbox2_output)
          

          
demo.launch(debug=True, share=True)

picture.image

Step4 ：我们在了解基于的操作之后，我们以一个GPT-2模型为例说明算法模型是如何部署的


          
# Let’s now build a simple interface that allows you to demo a text-generation model like GPT-2.
          
from transformers import pipeline
          

          
model = pipeline("text-generation")
          

          
def predict(prompt):
          
    completion = model(prompt)[0]["generated_text"]
          
    return completion

运行上述代码，会下载GPT-2模型

Step5 ：我们先来看一下GPT-2模型的预测结果


          
# Predict prompt based on GPT-2
          
predict("My favorite programming language is")

My favorite programming language is Python. It comes in both standard and custom forms, and each supports different types of manipulation. Here's a look at three of the most popular forms in Python, but you're welcome to share your own.

What

Step6 ：我们使用gradio部署模型来进行预测，由于GPT-2模型的生成是随机的，因此两次结果不一定会完全一致


          
# Using gradio API to visual the GPT-2 model
          
import gradio as gr
          

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

picture.image

gradio的主要API介绍：

Interface(fn, inputs, outputs, ...)

These parameters are:

fn: the prediction function that is wrapped by the Gradio interface. This function can take one or more parameters and return one or more values
inputs: the input component type(s). Gradio provides many pre-built components such as"image" or "mic".
outputs: the output component type(s). Again, Gradio provides many pre-built components e.g. "image" or "label".

You can customize the behavior of launch() through different parameters:

inline - whether to display the interface inline on Python notebooks.
inbrowser - whether to automatically launch the interface in a new tab on the default browser.
share - whether to create a publicly shareable link from your computer for the interface. Kind of like a Google Drive link!

完整的API介绍，请参考：https://gradio.app/docs/