You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Looks like it timeouts for bigger models, e.g. flux1-dev-fp8.safetensors while the original example works correctly. The modified workflow however works perfectly in the UI.
Super confused, would love to get some help.
The text was updated successfully, but these errors were encountered:
thanks for the detailed log messages. i've also heard of this happening to some others in our Slack channel. my hunch is that something buggy is happening with comfy run --workflow from comfy-cli where the websocket connection with the running comfyui server is getting lost. some users have talked about workarounds where they avoid websocket entirely and just trigger workflow runs / poll for requests using http. in any case, i'm also going to file an issue with comfy-cli if they have other insight.
While running the comfyapp.py and running the comfyclient.py locally - I'm getting errors while the same workflow works in the UI.
After running with verbose on, we do see that comfy is indeed running.
but errors out randomly after.
Looks like it timeouts for bigger models, e.g.
flux1-dev-fp8.safetensors
while the original example works correctly. The modified workflow however works perfectly in the UI.Super confused, would love to get some help.
The text was updated successfully, but these errors were encountered: