search

usefulsensors / openai-whisper

Robust Speech Recognition via Large-Scale Weak Supervision
MIT License
62 stars 24 forks source link

got "could not determine output format " error when audio file size small than 1MB #47

Closed jayyang-zigbang closed 1 year ago

jayyang-zigbang commented 1 year ago

Hi, i implemented a REST API using faster whisper.But i got a error when i receive audio file though "formdata" and put the audio data as binaryIO type for inferance. I found this issue only appear when audio file size small than 1MB. The big audio file works well.