使用 Hugging Face 的 Docker Spaces 建立機器學習應用程式

摘要

本文探討如何使用 Hugging Face 的 Docker Spaces 建立機器學習應用程式，對於希望提升開發效率與模型性能的讀者而言，這是一個不可錯過的重要主題。 歸納要點:

• Hugging Face 最新釋出的 LLM 訓練範本顯著提升了自然語言理解(NLU)和生成(NLG)的效能,特別在對話生成和摘要提取方面表現卓越。
• 結合 Docker Spaces 和 FastAPI,開發者可以輕鬆建立與最佳化容器,大幅縮短部署時間並降低資源消耗。
• Hugging Face 提供 CI/CD 功能,自動化端到端構建過程,使得機器學習模型的監控及快速部署更加高效。

我們在研究許多文章後,彙整重點如下

自然語言處理新篇章：大型語言模型與效能最佳化

#save it as app.py from fastapi import FastAPI from fastapi.middleware.cors import CORSMiddleware from fastapi.responses import RedirectResponse from transformers import pipeline  # Create a new FastAPI app instance app = FastAPI()  # Add CORS middleware app.add_middleware(     CORSMiddleware,     allow_origins=["*"],  # Allow all origins     allow_credentials=True,     allow_methods=["*"],  # Allow all methods     allow_headers=["*"],  # Allow all headers )  # Initialize the text generation pipeline pipe = pipeline("text2text-generation", model="google/flan-t5-small")  # Define a root path to verify the server is running @app.get("/", include_in_schema=False) def read_root():     return RedirectResponse(url="/docs")  # Define a function to handle the GET request at `/generate` @app.get("/generate") def generate(text: str):     """     Using the text2text-generation pipeline from `transformers`, generate text     from the given input text. The model used is `google/flan-t5-small`.     """     # Use the pipeline to generate text from the given input text     output = pipe(text)          # Return the generated text in a JSON response     return {"output": output[0]["generated_text"]}

#save it as requirements.txt fastapi uvicorn transformers torch

#save it as Dockerfile # Use the official Python 3.9 image FROM python:3.9  # Set the working directory to /code WORKDIR /code  # Copy the requirements.txt file into the container COPY requirements.txt .  # Install the dependencies RUN pip install --no-cache-dir --upgrade -r requirements.txt  # Set up a new user named "user" with user ID 1000 RUN useradd -m -u 1000 user  # Switch to the "user" user USER user  # Set home to the user's home directory ENV HOME=/home/user \     PATH=/home/user/.local/bin:$PATH  # Set the working directory to the user's home directory WORKDIR $HOME/app  # Copy the current directory contents into the container at $HOME/app setting the owner to the user COPY --chown=user . $HOME/app  # Expose the port FastAPI will run on EXPOSE 7860  # Start the FastAPI app on port 7860, the default port expected by Spaces CMD ["uvicorn", "app:app", "--host", "0.0.0.0", "--port", "7860"]

建立高效且可擴充套件的 FastAPI Docker 容器

提交檔案並監控構建狀態

參考來源

Taiwan-LLaMa Service: 使用FastAPI 部署成一個服務

PrimeHub — Our end-to-end MLOps platform is the entire machine learning lifecycle in one platform. ... Deploy Your Own API using FastAPI + Docker ...

來源： Infuseai.io

部署Machine Learning模型-使用Python,Pipeline, FastAPI,Docker, ...

了解從機器學習模型Machine Learning Model開發到部署的整個過程 · 了解如何用FastAPI開發高效的Web API · 了解如何使用Docker封裝FastAPI · 了解如何將封裝的API放到AWS EC2 ...

來源： Udemy

如何使用Streamlit 和FastAPI 构建即时机器学习Web 应用程序

完成所有数据分析、培训冠军模型并打包机器学习模型后，下一步是创建两个专用服务： 1 ） FastAPI 后端和2 ） Streamlit 前端。然后，这两个服务可以部署 ...

來源： NVIDIA Developer

Build Machine Learning Apps with Hugging Face

Using Spaces makes it easy to create and deploy ML-powered applications and demos in minutes. Recently, the Hugging Face team added support for ...

來源： Docker

Building a Machine Learning API with Python, FastAPI and Docker

Learn how to deploy a machine learning model with FastAPI and Docker.

來源： Medium

Index - FastAPI

FastAPI 是一個現代、快速（高效能）的web 框架，用於Python 並採用標準Python 型別提示。 主要特點包含： ... * 基於內部開發團隊在建立生產應用程式時的測試預估。

來源： FastAPI

tiangolo/full-stack-fastapi-template

Full stack, modern web application template. Using FastAPI, React, SQLModel, PostgreSQL, Docker, GitHub Actions, automatic HTTPS and more. License. MIT license.

來源： GitHub

Deploying Your FastAPI Applications on Huggingface Via Docker

Step 1: Create a new Docker Space. image/png. Next, you can ... TextGen/router.py - A Python script where we will write our FastAPI app.

來源： Hugging Face