https://github.com/voidful/hubert-pretrain
using huggingface trainer to pre-train hubert
https://github.com/voidful/hubert-pretrain
Last synced: 9 months ago
JSON representation
using huggingface trainer to pre-train hubert
- Host: GitHub
- URL: https://github.com/voidful/hubert-pretrain
- Owner: voidful
- Created: 2023-07-07T07:09:38.000Z (almost 3 years ago)
- Default Branch: main
- Last Pushed: 2023-10-26T13:27:01.000Z (over 2 years ago)
- Last Synced: 2024-12-31T00:17:59.174Z (over 1 year ago)
- Language: Python
- Size: 8.79 KB
- Stars: 2
- Watchers: 2
- Forks: 0
- Open Issues: 1
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
```shell
WANDB_MODE=disabled python train.py --tokenize_config facebook/hubert-large-ls960-ft --model_config \
ntu-spml/distilhubert --group_by_length --train_set hf-internal-testing/librispeech_asr_dummy --train_split validation \
--train_subset clean --test_split validation --test_subset clean --learning_rate 0.0003 --batch 30 --logging_steps 10 \
--eval_steps 60 --epoch 150 --use_auth_token True --output_dir ./model_test --overwrite_output_dir --batch 1 \
--vocab_size 500
```