Download baselines/Search-R1/scripts/download.py from ryan-superman/selfevo-a100-evacuation: direct link, hf CLI and curl.
- Browser
- Download file 834 Bytes
-
https://huggingface.co/ryan-superman/selfevo-a100-evacuation/resolve/main/baselines/Search-R1/scripts/download.py
- Command line
-
hf download hf://ryan-superman/selfevo-a100-evacuation/baselines/Search-R1/scripts/download.py
-
curl -L -o download.py https://huggingface.co/ryan-superman/selfevo-a100-evacuation/resolve/main/baselines/Search-R1/scripts/download.py
834 Bytes
| import argparse | |
| from huggingface_hub import hf_hub_download | |
| parser = argparse.ArgumentParser(description="Download files from a Hugging Face dataset repository.") | |
| parser.add_argument("--repo_id", type=str, default="PeterJinGo/wiki-18-e5-index", help="Hugging Face repository ID") | |
| parser.add_argument("--save_path", type=str, required=True, help="Local directory to save files") | |
| args = parser.parse_args() | |
| repo_id = "PeterJinGo/wiki-18-e5-index" | |
| for file in ["part_aa", "part_ab"]: | |
| hf_hub_download( | |
| repo_id=repo_id, | |
| filename=file, # e.g., "e5_Flat.index" | |
| repo_type="dataset", | |
| local_dir=args.save_path, | |
| ) | |
| repo_id = "PeterJinGo/wiki-18-corpus" | |
| hf_hub_download( | |
| repo_id=repo_id, | |
| filename="wiki-18.jsonl.gz", | |
| repo_type="dataset", | |
| local_dir=args.save_path, | |
| ) | |