Rumeysakeskin/ASR-Quantization
Post-training quantization on Nvidia Nemo ASR model
This project helps machine learning engineers and MLOps specialists speed up speech recognition by optimizing existing Nvidia NeMo ASR models. It takes a pre-trained ASR model and converts its internal computations to a lower precision format, resulting in a faster, more memory-efficient model for deployment on CPU devices. This is ideal for those managing the inference side of speech recognition systems.
No commits in the last 6 months.
Use this if you need to deploy an Nvidia NeMo ASR model for faster, more efficient speech recognition inference on a CPU device without retraining the model.
Not ideal if you are still in the training phase or if your primary deployment target is a GPU.
Stars
9
Forks
—
Language
Jupyter Notebook
License
—
Category
Last pushed
Aug 23, 2023
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/voice-ai/Rumeysakeskin/ASR-Quantization"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Featured in
Higher-rated alternatives
Uberi/speech_recognition
Speech recognition module for Python, supporting several engines and APIs, online and offline.
cmusphinx/pocketsphinx
A small speech recognizer
tensorflow/lingvo
Lingvo
modelscope/FunASR
A Fundamental End-to-End Speech Recognition Toolkit and Open Source SOTA Pretrained Models,...
PyThaiNLP/pythaiasr
Python Thai Automatic Speech Recognition