|
Download README.md from gemni16/processed3: direct link, hf CLI and curl.
- Browser
- Download file 1.02 kB
-
https://huggingface.co/datasets/gemni16/processed3/resolve/main/README.md
- Command line
-
hf download hf://datasets/gemni16/processed3/README.md
-
curl -L -o README.md https://huggingface.co/datasets/gemni16/processed3/resolve/main/README.md
1.02 kB
| pretty_name: processed3 | |
| # AgiBotWorld2025 condition data (processed3) | |
| This public dataset contains the existing `absolute_head_eef_state`, `action_map`, | |
| `action`, and `camera` directories from the audited `task_*` trees in | |
| `AgiBotWorld2025`. Archive members start with `AgiBotWorld2025/task_*`, so | |
| extracting at the destination preserves the dataset-relative tree. | |
| - Selected target directories: 672 | |
| - Regular files: 53376 | |
| - Uncompressed logical bytes: 116565618537 | |
| - Compression: one checksummed tar stream compressed with zstd and split into | |
| ordered 5 GiB byte parts | |
| Verify every downloaded part with `sha256sum -c SHA256SUMS`. Concatenate the | |
| paths in `PARTS` in exactly that order, pipe through `tools/bin/zstd -d -c`, | |
| and extract with GNU tar. Individual parts are not independently extractable. | |
| `manifest.json`, `inventory.jsonl`, and `missing_data.json` record the exact | |
| selection and source-side audit. The source tree was read only; all build | |
| artifacts were written outside `/mnt/cpfs/yangshuai`. | |