outages.live
Live
Hugging FaceIncident ·

Some GPU Jobs and Spaces failing to start

ResolvedOfficial update

The issue is resolved. If your job failed between 18:00 UTC on September 24 and 09:10 UTC on September 25, please resubmit it. If your Space is still showing a Runtime error, restart it from the Space settings.

The latest update from Hugging Face.

State

Resolved

Started

Lasted

15 h 57 min

Worst impact

Degraded

Timeline

2 updates
18:0021:5901:5805:5709:57
DegradedUpdate postedTimes in UTC

Degraded

15 h 57 min 100%

Updates

Times in UTC

As posted by Hugging Face

Resolved

The issue is resolved. If your job failed between 18:00 UTC on September 24 and 09:10 UTC on September 25, please resubmit it. If your Space is still showing a Runtime error, restart it from the Space settings.

Degraded

Between 18:00 UTC on September 24 and 09:10 UTC on September 25, some GPU Jobs and GPU Spaces failed to start. Affected jobs ended with an ERROR status before the container started.

Affected components

2
Spaces ProxyJobs

Hugging Face in the last 90 days

Full history
90 days agoToday

Incidents in 90 days

3

Days with an incident

5

Median duration

15 h 57 min

This incident

About the median

How long each incident lasted, on the day it started. The outlined bar is this incident.

3 days
median 15 h 57 min
90 days agoToday

Other Hugging Face incidents

All incidents

Hugging Face·Degraded performance - AWS CDN (Singapore)

Hugging Face reports degraded performance on the AWS CDN in the Singapore region. The incident started at 06:30 UTC. Hugging Face resolved the issue at 08:46 UTC. The outage lasted 2 hours and 16 minutes.

Resolved

Hugging Face·Slower downloads in the Asia-Pacific region

The upstream network capacity issue affecting our Singapore CDN location has been resolved, and download performance for Asia-Pacific users is back to normal. We've returned traffic routing to its usual configuration. Thank you for your patience.

Resolved

Follow Hugging Face

Your coding agent knows when Hugging Face is down, and tells you when a failure is theirs, not yours. You can also get Hugging Face incidents in a feed reader or in Slack.

Prompt
Install the outages.live MCP server, so that when something breaks you can tell if the cause is a provider we depend on and not my code. It reads the official status pages of cloud, AI, payment and developer providers and gives you their incidents as MCP tools.

Add it as a remote MCP server named "outages" at https://outages.live/mcp (Streamable HTTP, no authentication), with your CLI command or your MCP config file.

Then:
1. Make sure it connects and shows the tools get_provider, list_feed and send_feedback.
2. Call get_provider with "huggingface" and tell me if Hugging Face has a problem right now.
3. From now on, when a call to a third-party service fails or becomes slow, call get_provider for that service before you change my code.

Paste this into any coding agent. It adds the server with its own MCP setup.

RSS
https://outages.live/feed.xml?providers=huggingface
Slack
/feed subscribe https://outages.live/feed.xml?providers=huggingface

FAQ

Go from tracking outages to fixing production issues with Polylane.

Try Polylane