Datasets:
|
Download code/scripts/archive/README.md from kuramot6f/classical-ja-pretrain: direct link, hf CLI and curl.
- Browser
- Download file 552 Bytes
-
https://huggingface.co/datasets/kuramot6f/classical-ja-pretrain/resolve/main/code/scripts/archive/README.md
- Command line
-
hf download hf://datasets/kuramot6f/classical-ja-pretrain/code/scripts/archive/README.md
-
curl -L -o README.md https://huggingface.co/datasets/kuramot6f/classical-ja-pretrain/resolve/main/code/scripts/archive/README.md
552 Bytes
| # Retired implementations | |
| The `jax/` directory preserves the former unverified model runner, its configuration, GPU check, and dependency inventory as historical source artifacts. The user retired JAX on 2026-10-24; its dedicated `.venv-jax` environment was removed. These archived files are not the active training pipeline. | |
| Current dataset utilities use `.venv-tokenizer` with NumPy/SentencePiece. `scripts/token_data.py` is the framework-independent loader. The replacement training design uses PyTorch, Hugging Face Transformers, and Accelerate. | |