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
It stucked after loading the model.
Here is the console log captured from iphone:
[Error] Error while parsing the 'sandbox' attribute: 'allow-downloads' is an invalid sandbox flag. (whisper-web, line 225)
[Info] Successfully preconnected to https://fonts.gstatic.com/
[Error] Failed to load resource: the server responded with a status of 406 () (sse, line 0)
[Error] Failed to fetch Space status via SSE: – "Failed to initialize stream (code 406)"
onStateChange (SpaceHeader-BJTNEU2I.js:51:6726)
onerror (index.js:9:2862)
(anonymous function) (index.js:9:1545)
[Error] SSE Stream ended with error: – Error: Failed to initialize stream (code 406)
Error: Failed to initialize stream (code 406)
error
[Error] Unhandled Promise Rejection: Error: Unsupported model type: whisper
(anonymous function) (worker-8c8208dc.js:1790:179608)
[Error] Cache API operation failed: Quota exceeded
[Warning] Unable to add response to browser cache: QuotaExceededError: Quota exceeded. (worker-8c8208dc.js, line 1790)
@xenova please help to investigate.
Is this related with the JS engine in IOS causing the automatic pipeline and models having different behavior than chrome running in v8?
System Info
IOS safari
Transformer.js 2.7
Environment/Platform
Description
Whisper web demo here is not working for ios safari.
https://huggingface.co/spaces/Xenova/whisper-web
It stucked after loading the model.
Here is the console log captured from iphone:
Reproduction
open the demo site: https://huggingface.co/spaces/Xenova/whisper-web
choose whisper-base.en model
choose audio from the default url, click on transcribe button
The text was updated successfully, but these errors were encountered: