shhossain / banglaspeech2text

BanglaSpeech2Text: An open-source offline speech-to-text package for Bangla language. Fine-tuned on the latest whisper speech to text model for optimal performance.

Hi 👋, I'm sifat

• I’m currently exploring Transformers

• How to reach me [email protected]

Blogs posts

• The Evolution of Sorting Algorithms Over the Years (Bubble sort to AI-driven sort)
• The Evolution of Sorting Algorithms Over the Years (Bubble sort to AI-driven sort)
• Face Recognition on a Large Collection of Faces with Python
• How to Perform Face Recognition on a Large Collection of Faces with Python

Connect with me:

Languages and Tools:

banglaspeech2text's Issues

It is transcribing only few words with warning generation\utils.py:1353: UserWarning: Using max_length's default (448)

At first, thanks a lot for creating this awesome library.

I am getting the following warning:

\generation\utils.py:1353: UserWarning: Using max_length's default (448) to control the generation length. This behaviour is deprecated and will be removed from the config in v5 of Transformers -- we recommend using max_new_tokens to control the maximum length of the generation.

And it is transcribing only first few words. How can I transcribe the whole file with 5 minutes of audio?
Note: I am using windows

Is there a command line tool?

At first, thanks a lot for creating this nice lib and making it open source.
Is there any way to run in from cmd?
Thanks

Can we test?

Hi, the Colab notebook is not available. Can we test?