DeepSpeed Huggingface Automatic Speech Recognition Examples

October 13, 2022 ยท View on GitHub

Setup

Python dependencies:

pip install -r requirements.txt

For the test-wav2vec.py speech model example, you may also need to install the libsndfile1-dev generic library:

sudo apt-get install libsndfile1-dev

Usage

Examples can be run as follows:

deepspeed --num_gpus [number of GPUs] test-[model].py

Example Output

Command:

deepspeed --num_gpus 1 test-wav2vec2.py

Output:

WER: 0.03383673158855752