---
license: apache-2.0
base_model: yunyangx/efficient-track-anything
tags:
  - mirror
  - cuvis-ai
  - efficienttam
  - video-object-segmentation
---

# EfficientTAM checkpoints (mirror)

Byte-identical mirror of [`yunyangx/efficient-track-anything`](https://huggingface.co/yunyangx/efficient-track-anything) at revision `9bdd8ab585b19ef95f9c9ed847ac9478301890b4` (code and licence: <https://github.com/yformer/EfficientTAM>) for the
[Cuvis.AI](https://github.com/cubert-hyperspectral/cuvis-ai) plugins. No fine-tuning,
conversion or re-serialization: the files below are the upstream release as published,
with the upstream licence text in `LICENSE`.

## Why this mirror exists

Cuvis.AI provisions model weights once into a shared cache (`download-model download
<name>` from `cuvis-ai-core`) and then runs its pipelines in an offline, token-free
runtime that only reads that cache. Hosting the exact upstream files under the
`cubert-gmbh` organisation makes that provisioning reproducible (commit-pinned and
sha256-verified) and takes the per-user Hugging Face account and token out of the
users' path.

## Files and provenance

| File | Upstream source | Upstream revision / id | sha256 | Size |
|---|---|---|---|---|
| `efficienttam_s.pt` | `yunyangx/efficient-track-anything` / `efficienttam_s.pt` | `9bdd8ab585b19ef95f9c9ed847ac9478301890b4` | `2b572be30d9e96ee29c8d785fe157c6b079ede7d56fbc8a3671d4120e63c89cd` | 136.4 MB |
| `efficienttam_ti.pt` | `yunyangx/efficient-track-anything` / `efficienttam_ti.pt` | `9bdd8ab585b19ef95f9c9ed847ac9478301890b4` | `acbb17b28cca1f860acee09c9ecb6efdb732080dc7a85a07292c31813175fa7d` | 71.6 MB |
| `efficienttam_s_512x512.pt` | `yunyangx/efficient-track-anything` / `efficienttam_s_512x512.pt` | `9bdd8ab585b19ef95f9c9ed847ac9478301890b4` | `67b5840012737ed2c94a4cb8787c5c1b27b3a946045d2e602e65ef77230b6085` | 136.4 MB |
| `efficienttam_ti_512x512.pt` | `yunyangx/efficient-track-anything` / `efficienttam_ti_512x512.pt` | `9bdd8ab585b19ef95f9c9ed847ac9478301890b4` | `7d4d652a465f0081391050932f45d8a66768ccc99c8ea393ce8a5927e83f3b9b` | 71.6 MB |
| `LICENSE` | <https://raw.githubusercontent.com/yformer/EfficientTAM/main/LICENSE> | n/a (URL; sha256 pinned) | `c71d239df91726fc519c6eb72d318ec65820627232b2f796219e87dcf35d0ab4` | 11.4 kB |

Mirrored on 2026-09-04 by Cubert GmbH from the sources above. The sha256 values are the
upstream values; `tools/mirror_weights.py check` in `cuvis-ai-core` re-verifies this
repository against them and against the upstream licence text.

## Licence

Apache License 2.0. The `LICENSE` file is the upstream licence text taken verbatim
from <https://github.com/yformer/EfficientTAM> (the upstream Hugging Face repository carries the `apache-2.0` tag but ships no LICENSE file) at the mirror date. Copyright remains with the upstream authors; Cubert
GmbH claims no rights in these files.

## Usage with Cuvis.AI

```text
uv run download-model download efficienttam_s
uv run download-model download efficienttam_ti
uv run download-model download efficienttam_s_512x512
uv run download-model download efficienttam_ti_512x512
```

`download-model` (from `cuvis-ai-core`) provisions the file into the shared Hugging
Face cache and verifies its sha256. From cuvis-ai-core 0.16.0 on,
`ModelWeights.resolve("<name>")` returns the cached path and the `cuvis-ai-rtsam2` nodes
resolve their weights through it, so a pipeline needs no manual step once the
weights are provisioned; earlier plugin releases still fetch from their original
upstream sources.