Instructions to use mohitsha/whisper-tiny-smooth-quant with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use mohitsha/whisper-tiny-smooth-quant with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("automatic-speech-recognition", model="mohitsha/whisper-tiny-smooth-quant")# Load model directly from transformers import AutoProcessor, AutoModelForSpeechSeq2Seq processor = AutoProcessor.from_pretrained("mohitsha/whisper-tiny-smooth-quant") model = AutoModelForSpeechSeq2Seq.from_pretrained("mohitsha/whisper-tiny-smooth-quant") - Notebooks
- Google Colab
- Kaggle
open-sourced codes for the edited framework of smoothquant for whisper
#1
by KukiFan - opened
Hello mohitsha,
I am a student currently learning how to use SmoothQuant to quantize Whisper. I noticed that SmoothQuant cannot be directly applied to Whisper out of the box. Would it be possible for you to open-source your code for quantizing Whisper-tiny with SmoothQuant?
Your work appears to be the only successful implementation of this so far. I really admire your ingenuity and would greatly appreciate your help!
Thank you so much!