metadata
license: bsd-3-clause
tags:
- mirror
- xai
- video
- action-recognition
- kinetics-400
R3D-18 Kinetics-400 (torchvision) — xaitalk mirror
This is a bit-identical mirror of the canonical artifact from PyTorch / torchvision.
The mirror exists only as a resilience fallback for the xaitalk library — the upstream remains authoritative. All credit and licensing for the model belong to the original authors.
Attribution
| Field | Value |
|---|---|
| Original authors | PyTorch / torchvision |
| Upstream (authoritative) | https://download.pytorch.org/models/r3d_18-b3b3357e.pth |
| Source repo | https://github.com/pytorch/vision |
| Paper | https://arxiv.org/abs/1711.11248 (Tran et al. 2018, A Closer Look at Spatiotemporal Convolutions) |
| License | bsd-3-clause (inherited from upstream — please respect upstream's terms) |
| Mirror file | r3d_18-b3b3357e.pth |
| SHA-256 | b3b3357ead25631ec9c57362ff2128a92d0427e01e2cd184951a44380c3f2e9d |
| Size | 133,546,016 bytes (127.4 MB) |
How xaitalk loads this file
from xaitalk.hub import ensure_model
weights_path = ensure_model("r3d18-kinetics400")
# Tries the canonical upstream first; falls back to this xaitalk mirror
# automatically if upstream is unreachable.
Why mirror?
xaitalk's research-grade reproducibility claim relies on every weight file
being recoverable years from now. We mirror artifacts ≤ 2.5 GB under
xaitalk/*-mirror so the pipeline survives upstream URL changes, repo
renames, or deletions. Bit-level parity with the canonical is asserted in
CI via python -m xaitalk.hub verify-mirrors.
Citation
If you use this model, please cite the original paper (not the mirror):
https://arxiv.org/abs/1711.11248 (Tran et al. 2018, A Closer Look at Spatiotemporal Convolutions)