This training code is based on transformers/examples/pytorch/speech-pretraining in https://github.com/huggingface/transformers
check.py
checks a target model and datasetextract_result.py
createsresult.pkl
, storeing embedding vectors from all audio samples in a target datasetplot_result.py
: plots embeding space fromresult.pkl
model_push_to_hub.py
: pushes model to huggingface
sh run_birddb.sh
This script performs speech-pretraining for bird songs
using run_wav2vec2_pretraining_no_trainer.py
.