Melih Yılmaz ÖGÜTCEN

You can use this full script for testing. I uploaded an English sound to GitHub. By using this, you can try it too. ``` from six.moves.urllib.request import urlopen import io...

I fixed my problem by updating `generation_config.json`. Thanks!

Oops! I have tried different sounds with the new config. And rarely, I got this error again on some sounds. ``` Traceback (most recent call last): File "/SpeechToText/whisper_trials.py", line 63,...

Sorry, I couldn't share these files due to privacy, but I can send the `previous_tokens`. I added print function here. https://github.com/huggingface/transformers/blob/main/src/transformers/models/whisper/tokenization_whisper.py#:~:text=current_tokens%20%3D%20%5B%5D-,if%20previous_tokens%3A,-if%20return_timestamps%3A Is it correct? ``` Previous tokens: [[16729, 44999, 39196,...

Yes, it is Turkish and I used `whisper-large-v2.` I used the same script as above, I just used "" language and I changed `generation_config.json` as you said.

@themanyone I tried your script with the whisper-large-v2 model, and the output is like below. ``` 2023-04-25 14:58:25.859863: I external/xla/xla/service/service.cc:168] XLA service 0xa2ff910 initialized for platform Interpreter (this does not...

Here, I used recommended parameters (https://huggingface.co/blog/asr-chunking https://colab.research.google.com/drive/1rS1L4YSJqKUH_3YxIQHBI982zso23wor?usp=sharing#scrollTo=Mh_e6rV62QUM) @themanyone In these two codes parameters, sound data, and environments are the same. Still, there is no acceleration. @sanchit-gandhi

Yes, in the first transcription, it is highly slower, but in the second time yes it is much better than the first transcription but still, it is not faster than...