Download source/src/vimeml/tokenizer/native_train.py from Voltline/vimeml-tiny-ja-v2.1: direct link, hf CLI and curl.
- Browser
- Download file 387 Bytes
-
https://huggingface.co/Voltline/vimeml-tiny-ja-v2.1/resolve/main/source/src/vimeml/tokenizer/native_train.py
- Command line
-
hf download hf://Voltline/vimeml-tiny-ja-v2.1/source/src/vimeml/tokenizer/native_train.py
-
curl -L -o native_train.py https://huggingface.co/Voltline/vimeml-tiny-ja-v2.1/resolve/main/source/src/vimeml/tokenizer/native_train.py
387 Bytes
| """Child process capturing SentencePiece's native C++ training diagnostics.""" | |
| import json | |
| import sys | |
| from pathlib import Path | |
| def main(): | |
| import sentencepiece as spm | |
| request = json.loads(Path(sys.argv[1]).read_text(encoding='utf-8')) | |
| spm.set_random_generator_seed(request['seed']) | |
| spm.SentencePieceTrainer.train(**request['options']) | |
| if __name__=='__main__': main() | |