/wav2vec2-sprint

Primary LanguageJupyter NotebookMIT LicenseMIT

This repo is deprecated in favor of https://github.com/jonatasgrosman/huggingsound

Wav2Vec Trainer

This repository is based on https://github.com/jqueguiner/wav2vec2-sprint

Building docker image

Dockerhub available at https://hub.docker.com/r/patilsuraj/hf-wav2vec

to build the docker :

$ docker build -t hf-wav2vec-sprint -f Dockerfile .

to push it to dockerhub First create a repository on dockerhub

$ docker tag hf-wav2vec-sprint your-dockerhub-user/hf-wav2vec-sprint

to push it to dockerhub

$ docker push your-dockerhub-user/hf-wav2vec-sprint

Running WandB sweep

Initialize your sweep from any machine...

$ export WANDB_API_KEY=YOUR_WANDB_API_KEY
$ export WANDB_ENTITY=YOUR_WANDB_ENTITY
$ export WANDB_PROJECT=YOUR_WANDB_PROJECT

$ wandb sweep sweep.yaml

... the execution above will give you a sweep id, save it and on the training machine run:

$ export WANDB_API_KEY=YOUR_WANDB_API_KEY
$ export WANDB_ENTITY=YOUR_WANDB_ENTITY
$ export WANDB_PROJECT=YOUR_WANDB_PROJECT

$ wandb agent YOUR_SWEEP_ID

Uploading model to HF

You need to upload the following files to the HF repository

  • preprocessor_config.json
  • special_tokens_map.json
  • tokenizer_config.json
  • vocab.json
  • config.json
  • pytorch_model.bin
  • README.md (create this file based on the MODEL_CARD.md)
$ git config --global user.email "email@example.com"

$ git config --global user.name "Your name"

$ transformers-cli login

$ transformers-cli repo create your-model-name

$ git clone https://username:password_or_token@huggingface.co/username/your-model-name

$ git add .

$ git commit -m "Initial commit"

$ git push

Troubleshooting

  • audioread.exceptions.NoBackendError: $ sudo apt-get install ffmpeg sox libsox-fmt-mp3

Finetuned models

Wav2Vec2-XLSR-53