5386b62dc4
chore(ml): allow insightface 1.x ( #28595 )
...
* chore(ml): allow insightface 1.x
The new insightface 1.0 release appears to have no breaking code changes nor relevant license changes ([before](https://github.com/deepinsight/insightface/blob/2a78baec428354883e0cda39c54b555a5ed8358a/README.md ), [after](https://github.com/deepinsight/insightface/blob/70f3269ea628d0658c5723976944c9de414e96f8/README.md ), c.f. https://github.com/immich-app/immich/blob/fd7ddfef54cdf2b6256c4fc08bc5ff3f86176775/machine-learning/README.md ), and it works on my machine.
* Update uv.lock
* please excuse my incompetence
2026-05-25 12:32:50 -04:00
a5a7380a26
feat: use lockfile for mise tools ( #28503 )
2026-05-20 11:37:33 +00:00
8c8dc9d32f
chore(ml)!: remove deprecated envs ( #28326 )
...
remove deprecated envs
2026-05-09 22:40:05 +00:00
25a6a38b30
chore: use mise ( #28298 )
2026-05-08 15:21:33 -04:00
bfd76570c5
chore(deps): update dependency python-multipart to v0.0.27 [security] ( #28286 )
...
Co-authored-by: renovate[bot] <29139614+renovate[bot]@users.noreply.github.com>
2026-05-07 15:22:03 -04:00
36caeb34ec
chore(ml)!: require numpy 2.4 ( #28158 )
...
require numpy 2.4
2026-05-07 19:07:39 +00:00
52b00b0bad
chore(ml): add mise checklist command ( #28267 )
...
* chore(ml): add mise checklist command
* don't depend tests on installing a cpu flavor
2026-05-07 12:28:24 -04:00
d93ab7707e
chore(ml): configure mise for machine-learning directory ( #25579 )
2026-05-06 11:03:54 -04:00
eca0e60db8
fix: librknnrt permissions in machine-learning ( #28216 )
...
fix librknnrt permissions in machine-learning
2026-05-03 23:39:27 +00:00
8cff5883b5
fix(ml): respect time zone for logs in cuda container ( #28155 )
2026-05-03 04:19:56 +00:00
5e89efba64
fix(ml): handle empty/corrupt images in face detection ( #27391 )
...
* fix(ml): handle empty/corrupt images in face detection
When a corrupt or degenerate image with zero-dimension (0 width or 0 height)
reaches the face detection pipeline, insightface's RetinaFace.detect() calls
cv2.resize() with a target size of 0, triggering an OpenCV assertion failure:
error: (-215:Assertion failed) inv_scale_x > 0 in function 'resize'
This crashes the ML worker and returns a 500 error to the server.
Add an early return in FaceDetector._predict() that checks for zero-dimension
images after decoding and returns empty detection results instead of passing
them to the insightface model.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com >
* fix(ml): move empty image validation to request level
Per review feedback, validate image dimensions in the predict endpoint
(returning 400) rather than in each model's _predict method. This
catches all zero-dimension images before they reach any model task.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com >
* fix(ml): resolve mypy strict type error in predict endpoint
Use intermediate `decoded` variable so mypy knows `.width` and `.height`
are accessed on `Image`, not on `Image | str`.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com >
---------
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com >
2026-04-27 11:14:34 -04:00
9263e2f2e1
feat(ml): update Intel graphics compiler and compute runtime ( #28076 )
...
feat(ml): update Intel graphics compiler and compute runtime to latest versions
2026-04-25 08:49:57 -04:00
a3ee615c5b
chore(ml): update huggingfacehub and pillow ( #27552 )
2026-04-24 19:44:01 -04:00
95fa8fbdab
chore(deps): update machine-learning ( #26970 )
...
Co-authored-by: renovate[bot] <29139614+renovate[bot]@users.noreply.github.com>
2026-04-16 11:32:08 -04:00
4ff6cca4da
fix(deps): update dependency pillow to >=12.2,<12.3 [security] ( #27773 )
...
Co-authored-by: renovate[bot] <29139614+renovate[bot]@users.noreply.github.com>
2026-04-16 09:13:39 -04:00
2b7ae4981f
chore(deps): update dependency pytest to v9.0.3 [security] ( #27777 )
...
Co-authored-by: renovate[bot] <29139614+renovate[bot]@users.noreply.github.com>
2026-04-16 09:12:49 -04:00
2ff9f95527
chore(deps): update dependency python-multipart to v0.0.26 [security] ( #27838 )
...
Co-authored-by: renovate[bot] <29139614+renovate[bot]@users.noreply.github.com>
2026-04-15 19:06:27 -04:00
95e57a24cb
chore: version v2.7.5
2026-04-13 14:27:31 +00:00
bee49cef02
chore: version v2.7.4
2026-04-10 16:32:26 +00:00
3254d31cd2
chore: version v2.7.3
2026-04-09 17:51:40 +00:00
6a63e814a5
chore: version v2.7.2
2026-04-07 20:58:38 +00:00
b03a649e74
chore: version v2.7.1
2026-04-07 20:22:28 +00:00
9ba9a22c40
fix(ml): downgrade numpy ( #27591 )
...
downgrade numpy
2026-04-07 15:57:42 -04:00
afec61addc
chore: version v2.7.0
2026-04-07 15:08:18 +00:00
a2ff075e9a
chore: version v2.6.3
2026-03-26 16:23:35 +00:00
ce9b32a61a
chore: version v2.6.2
2026-03-24 02:51:55 +00:00
e939fde6f1
chore: version v2.6.1
2026-03-19 17:56:42 +00:00
f413f5c692
chore: version v2.6.0
2026-03-18 20:37:39 +00:00
aaf34fa7d4
feat(ml): enable openvino for cpu ( #22948 )
...
* Enable OpenVINO CPU acceleration in Immich
* Remove unnecessary debug log
* Removing checking for device_ids for openvino since cpu will always be available
* Find OpenVINOExecutionProvider index instead of assuming index 0
* Fix openvino tests
* Fix failing test mock. OpenVINO expects provider options, but cuda provide doesn't so use that for mocked tests.
* Support empty provider options in OrtSessions in which case ONNXRuntime will use its own defaults
* Use OpenVINOExecutionProvider for test_sets_provider_options_kwarg
* fix mock
* simplify
* unused variable
---------
Co-authored-by: Aleksander <pejcic@adobe.com >
Co-authored-by: mertalev <101130780+mertalev@users.noreply.github.com >
2026-03-07 18:40:43 +00:00
35a521c6ec
fix(ml): batch size setting ( #26524 )
2026-03-05 12:01:47 -05:00
dc4da4b3d6
feat: update onnxruntime-openvino to 1.24.1 and intel drivers ( #26565 )
...
feat: update onnxruntime-openvino to 1.24.1 and intel drivers to the latest version
2026-02-27 16:35:29 -05:00
dd9046508d
feat: ROCm 7.2 and MIGraphX support ( #26178 )
2026-02-26 16:52:26 +00:00
dd97395f3a
chore(deps): update dependency gunicorn to v25 ( #26486 )
...
Co-authored-by: renovate[bot] <29139614+renovate[bot]@users.noreply.github.com>
2026-02-24 16:14:04 +00:00
7461479f60
chore(ml): remove unused dependency ftfy ( #25529 )
2026-02-19 22:58:25 +00:00
fea6e8d9f3
chore(deps): update dependency python-multipart to v0.0.22 [security] ( #25559 )
2026-02-12 09:32:59 -05:00
27ebbab1d9
fix(deps): update dependency pillow to v12 [security] ( #26142 )
2026-02-12 09:32:17 -05:00
4647ecf2ea
chore(deps): update machine-learning ( #25067 )
2026-02-12 09:31:13 -05:00
3be8e265cd
chore: version v2.5.6
2026-02-10 18:24:40 +00:00
937bef9a4d
fix: run rocm builds on pokedex ( #26062 )
...
* fix: run rocm builds on pokedex
* fix: --parallel 48 (#26065 )
2026-02-09 15:06:05 -05:00
ff9052f7f5
chore: version v2.5.5
2026-02-06 03:49:02 +00:00
8581b4f350
chore: version v2.5.4
2026-02-05 21:16:52 +00:00
38c1f0b1fd
chore: version v2.5.3
2026-02-03 18:14:21 +00:00
9f52d864cf
chore(ml): replace black with ruff format ( #25578 )
2026-02-02 09:02:06 -05:00
eadb2f89af
chore: version v2.5.2
2026-01-28 22:05:10 +00:00
41e2ed3754
chore: version v2.5.1
2026-01-27 23:10:13 +00:00
6fd3c9fffa
chore: version v2.5.0
2026-01-27 18:19:23 +00:00
fdff591a11
feat: update intel compute driver ( #25259 )
2026-01-16 14:42:55 +00:00
3321c1a9df
feat(ml): update ONNX Runtime, OpenVINO and ROCm stack ( #23458 )
2026-01-01 12:17:55 -05:00
e63e8e2517
chore(deps): update machine-learning ( #24610 )
...
Co-authored-by: renovate[bot] <29139614+renovate[bot]@users.noreply.github.com>
2025-12-23 03:12:13 +00:00
50d7956c07
chore: version v2.4.1
2025-12-19 15:03:03 +00:00
Aaron Liu
bo0tzz
Mert
Jason Rasmussen
renovate[bot]
Timon
David Allen
AyaanMAG
Yosi Taguri
Savely Krasovsky
github-actions
Aleksander Pejcic
Kishor Prins
dotlambda