I can’t load my new Space and one previous I redeployed. Tried to do restart, factory load, run with Docker and with Gradio. Once it loaded and started working but then with my next commit failed again.
OSError: We couldn't connect to 'https://huggingface.co' to load this file, couldn't find it in the cached files and it looks like laiyer/deberta-v3-base-injection-onnx is not the path to a directory containing a file named config.json.
Checkout your internet connection or see how to run the library in offline mode at 'https://huggingface.co/docs/transformers/installation#offline-mode'.
And this error
requests.exceptions.ConnectionError: (MaxRetryError('HTTPSConnectionPool(host=\'huggingface.co\', port=443): Max retries exceeded with url: /laiyer/deberta-v3-base-injection-onnx/resolve/main/config.json (Caused by NameResolutionError("<urllib3.connection.HTTPSConnection object at 0x7fe0cd3e90f0>: Failed to resolve \'huggingface.co\' ([Errno -3] Temporary failure in name resolution)"))'), '(Request ID: c0a3edd4-8f32-4c8b-92d6-506f025e16af)')
Hello there. So… I’m not the only one who experience that huge connection issue. But he way, before experiencing that issue, I was stuck between an infinite queue on free api inference. Each time I attempted a new inference, it just increased from 1 the queue. Even if I closed my connection (or even the page) to run my private space test in a new webbrowser tab.
To fix it, it seems duplicate the space which experiencing errors and waiting the build of the newer one is the key. But I clearly think that it is a connection glitch on build/inference-api of huggingface.co.
Edit : also I have a another space that won’t run, it is stuck at building page step. Think the error is the same : Connection issue from the Docker container to external adress (note : external or local such a model referenced on Huggingface.co).
Yes - mine is reporting an “unable to resolve host” on the first outcall (to get SD) - and after a handful of failed retries, the build crashes (crash is probably due to lack of error handling in the script, though).
My spaces aren’t building too, yesterday they did, now even after several factory rebuilds, restarts etc. they stuck in rebuilding.
Its error is the following: Launch timed out, space was not healthy after 30 min. Thats it.
Mine is a Streamlit though.
Now for my part, it is reworking as expected. It seems it was a huggingface.co build servers issues. Try to pause your space and unpause it (resume), from the Settings tab in the Space!