The below example is broken:
response = client.audio.transcriptions.create(
model: "whisper-1",
file: File.open(mp3_filepath.to_s, "rb"),
response_format: :verbose_json,
timestamp_granularities: [ :word, :segment ],
language: "en"
)
The response does not include the list of words and their timestamps. This is because timestamp_granularities is not getting properly encoded. I've created a working, corresponding fix to this issue here #249
The below example is broken:
The response does not include the list of words and their timestamps. This is because
timestamp_granularitiesis not getting properly encoded. I've created a working, corresponding fix to this issue here #249