Datasets:
Download scripts/build_envs.sh from divy123/sequential-router-v2: direct link, hf CLI and curl.
- Browser
- Download file 953 Bytes
-
https://huggingface.co/datasets/divy123/sequential-router-v2/resolve/main/scripts/build_envs.sh
- Command line
-
hf download hf://datasets/divy123/sequential-router-v2/scripts/build_envs.sh
-
curl -L -o build_envs.sh https://huggingface.co/datasets/divy123/sequential-router-v2/resolve/main/scripts/build_envs.sh
953 Bytes
| set -e | |
| export UV_CACHE_DIR=/mnt/localssd/uv_cache TMPDIR=/mnt/localssd/tmp HF_HOME=/mnt/localssd/hf_cache | |
| V=/mnt/localssd/venvs | |
| uv venv $V/main --python 3.12 --seed | |
| uv pip install --python $V/main/bin/python vllm | |
| uv pip install --python $V/main/bin/python pandas pyarrow datasets scikit-learn scipy matplotlib lightgbm peft sacrebleu rouge-score bert-score nltk langdetect absl-py immutabledict httpx pyyaml tqdm huggingface_hub math-verify | |
| $V/main/bin/python -c "import vllm, torch, transformers; print('main ok', vllm.__version__, torch.__version__, transformers.__version__, torch.cuda.device_count())" | |
| uv venv $V/comet --python 3.12 --seed | |
| uv pip install --python $V/comet/bin/python torch --index-url https://download.pytorch.org/whl/cu128 | |
| uv pip install --python $V/comet/bin/python unbabel-comet pandas pyarrow pyyaml sacrebleu | |
| $V/comet/bin/python -c "import comet, torch; print('comet ok', torch.__version__)" | |
| echo ENVS_DONE | |