Skip to main content

Upload files to a Hugging Face space using the CLI and Python, including remote file uploads

How to upload files to a Hugging Face space using the CLI and Python, including remote file uploads:

Uploading Files to Hugging Face Spaces

Hugging Face Spaces allow you to easily host models and apps. A key part of managing a Space is uploading files - like datasets, model files, images, etc. Here are some tips on how to upload files to your Space using the Hugging Face CLI and Python library.

Uploading Local Files

To upload a local file to your Space from the command line, use the huggingface-cli upload command:

huggingface-cli upload username/my-space ./local/file.txt

This will upload file.txt directly to the root of your Space. You can also specify a path within the Space:

huggingface-cli upload username/my-space ./local/data.csv data/data.csv

In Python, install the huggingface_hub library and use upload_file():

from huggingface_hub import login, upload_file

login(token="YOUR_TOKEN")

upload_file(
  path_or_fileobj="./local/image.png",
  path_in_repo="images/profile.png",
  repo_id="username/my-space"
)

You can upload entire folders using upload_folder() instead.

Uploading Remote Files

To upload a file from a URL to your Space, you can use the huggingface-cli with the --remote flag:

huggingface-cli upload username/my-space --remote https://example.com/data.csv

This will download the file and upload it. In Python, you can use the requests library:

import requests
from huggingface_hub import login, upload_file

url = "https://example.com/file.txt"

# Download file
response = requests.get(url)
contents = response.content

login(token="YOUR_TOKEN") 

upload_file(
  path_or_fileobj=contents,
  path_in_repo="remote_data.txt",
  repo_id="username/my-space"  
)

This way you can download a remote file and directly upload the contents to your Space in one step.

Summary

Key points:

  • Use huggingface-cli upload to upload local files and folders
  • Pass --remote to upload directly from a URL
  • In Python, use upload_file() and upload_folder() from huggingface_hub
  • Download remote files first before uploading the contents

This provides some handy code snippets to easily manage files in your Hugging Face Spaces!

Comments

Popular posts from this blog

Top Free APIs Every Developer Should Know About

Top Free APIs Every Developer Should Know About In the world of software development, APIs (Application Programming Interfaces) are essential for integrating various functionalities into applications. Here’s a curated list of top free APIs categorized by their functionality: 1. Weather APIs OpenWeatherMap API : Provides current weather data, forecasts, and historical weather data for any location. Weatherstack API : Offers real-time weather information, including forecasts and historical data. 2. Maps and Geolocation APIs Google Maps API : Enables integration of interactive maps, geocoding, and route optimization. Mapbox API : Provides customizable maps, navigation, and location search capabilities. 3. Finance and Stock Market APIs Alpha Vantage API : Offers real-time and historical equity and cryptocurrency data. Yahoo Finance API : Provides access to financial news, stock market data, and por...

Google Drive Proxy Video Player - Bypass Limits - JW Player - Embed drive videos

GooDrive :- https://goodrive.stream/ Google Drive Proxy Player #1 :- https://youtu.be/9VQK8W2iUkg Dev.to Article

Making AI Song Covers with RVC - Google Colab

Making AI Song Covers with RVC * Google Colab or Local Install These are the two main options for making AI song covers. You can run RVC on your computer if you have a PC with a decent NVIDIA graphics card (GPU), or you can run it for free through the Google Colab web page. Running Google Colab This is the recommended Google Colab for using voice models: https://colab.research.google.com/drive/1Gj6UTf2gicndUW_tVheVhTXIIYpFTYc7?usp=sharing After enough time, Google limits your GPU usage and you have to wait to use the GPU again. This will slow down your conversion speeds, but it will still be usable as long as you use ‘rmvpe’ mode (considered to be the general best mode, tied with mangio-crepe). ~3 minute song took 9 minutes for me without the GPU. Some people make alternate Google accounts to get around the GPU limits, or they pay for Colab Pro. Most commonly happens for people training their own voices since that requires a lot of GPU power. Running Locally Check...

Random Posts