Notifications
Clear all
Features
5
Posts
3
Users
0
Reactions
4
Views
Topic starter
Hi everyone.
I’m trying to make an API call in OpenAI Whisper Transcription. Specifically, the endpoint is /v1/audio/transcriptions. My objective is to obtain a transcription that includes timestamps, which I'm unable to achieve with the current Whisper Transcription module. Therefore, I want to replicate the API call and manually add the necessary parameters.
However, I’m encountering this error:
The file “value” parameter is directly sourced from a download file module, and the data is essentially an IMTBuffer.
Any suggestions on how to resolve this? Thank you.
Posted : 17/03/2024 10:57 am
Hello there!
It appears the JSON provided is not valid. You can utilize online tools to verify JSON validity.
Here's a helpful tool: https://jsonlint.com/
Should you need further assistance, feel free to contact us.
Visit us here
Posted : 18/03/2024 4:07 am
Topic starter
I suspect this is related to the {{data}} value.
However, the value is directly mapped from a download module and contains IMTBuffer binary data, so it ought to function correctly.
Posted : 18/03/2024 5:34 am
I'm encountering a similar situation and found this thread. My understanding is that this functionality might not be achievable at the moment. Encoding binary data directly into JSON, based on my findings, is unlikely to be successful. Additionally, the API call option doesn't seem to support a direct upload that would be included in the POST request. Therefore, I don't believe this can be accomplished currently.
However, if you've managed to make it work, I would greatly appreciate it if you could share your progress.
Posted : 02/04/2024 8:28 pm
Topic starter
The openAI callin.io API Call won’t be able to handle this as the encoded binary data is not possible in JSON.
A quick workaround was to use an HTTP module and set the content type to “form data” to allow mapping of the actual file.
Posted : 03/04/2024 6:54 am