Glenn Jocher
|
7e99804263
|
Cleanup argument handling in `Model` class (#4614)
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Sebastian Stapf <42514241+Wiqzard@users.noreply.github.com>
|
2 years ago |
|
Glenn Jocher
|
deac7575b1
|
Model coverage cleanup (#4585)
|
2 years ago |
|
Glenn Jocher
|
1db9afc2e5
|
Fix Windows non-UTF source filenames (#4524)
Co-authored-by: Kayzwer <68285002+Kayzwer@users.noreply.github.com>
|
2 years ago |
|
Glenn Jocher
|
6da8f7f51e
|
Add `utils.ops` and `nn.modules` to tests (#4484)
|
2 years ago |
|
Glenn Jocher
|
615ddc9d97
|
Tests and docstrings improvements (#4475)
|
2 years ago |
|
Glenn Jocher
|
87ce15d383
|
`ultralytics 8.0.158` add benchmarks to coverage (#4432)
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Yonghye Kwon <developer.0hye@gmail.com>
|
2 years ago |
|
Glenn Jocher
|
9d27e7ada4
|
New `ASSETS` and trackers GMC cleanup (#4425)
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
|
2 years ago |
|
Glenn Jocher
|
fb1ae9bfad
|
Fix `LoadStreams` final frame bug (#4387)
Co-authored-by: Nadim Bou Alwan <64587372+nadinator@users.noreply.github.com>
|
2 years ago |
|
Glenn Jocher
|
d47718c367
|
`ultralytics 8.0.154` add `freeze` training argument (#4329)
|
2 years ago |
|
Glenn Jocher
|
9f6d48d3cf
|
Improve tests coverage and speed (#4340)
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
|
2 years ago |
|
Glenn Jocher
|
28893d9d2d
|
`ultralytics 8.0.152` bump `torch>=1.8.0` (#4299)
|
2 years ago |
|
Glenn Jocher
|
7dfdb63cde
|
`ultralytics 8.0.150` new Apple `*.mlpackage` export format (#4043)
Co-authored-by: MLBoy_DaisukeMajima <rockyshikoku@gmail.com>
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Louis Lac <lac.louis5@gmail.com>
|
2 years ago |
|
Glenn Jocher
|
7565210484
|
`ultralytics 8.0.149` add Open Images V7 training (#4178)
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: AdiEcho <30563671+AdiEcho@users.noreply.github.com>
|
2 years ago |
|
Glenn Jocher
|
11d0488bf1
|
Update prediction `Results` docs (#4139)
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
|
2 years ago |
Sebastian Stapf
|
8870084645
|
Removed redundency in RTDETRDecoder (#4118)
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Glenn Jocher <glenn.jocher@ultralytics.com>
Co-authored-by: Laughing-q <1185102784@qq.com>
Co-authored-by: Laughing <61612323+Laughing-q@users.noreply.github.com>
|
2 years ago |
|
Glenn Jocher
|
379f31904a
|
Fix per-layer FLOPs profiling (#4048)
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
|
2 years ago |
Motoki Kimura
|
3c787eb080
|
Fix catastrophic accuracy degradation of TFLite static quantized integer models (#1695)
Co-authored-by: Glenn Jocher <glenn.jocher@ultralytics.com>
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
|
2 years ago |
Birk Stoveland
|
42afe772d5
|
Update tasks.py to allow custom RTEDTRDecoder arguments (#3872)
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Glenn Jocher <glenn.jocher@ultralytics.com>
|
2 years ago |
nihui
|
01dcd54b19
|
Add ncnn auto thread count (#3847)
Co-authored-by: Glenn Jocher <glenn.jocher@ultralytics.com>
|
2 years ago |
Colin Wong
|
b251b7415a
|
Fix tflite int8 scaling (#3837)
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Glenn Jocher <glenn.jocher@ultralytics.com>
|
2 years ago |
|
nihui
|
cf3c740fe7
|
update cli.md for export ncnn model (#3846)
|
2 years ago |
Laughing
|
620f3eb218
|
`ultralytics 8.0.136` refactor and simplify package (#3748)
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Glenn Jocher <glenn.jocher@ultralytics.com>
|
2 years ago |
|
Glenn Jocher
|
114470361e
|
Remove deprecated `AutoShape` class (#3743)
|
2 years ago |
|
Glenn Jocher
|
135a10f1fa
|
Avoid CUDA round-trip for relevant export formats (#3727)
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
|
2 years ago |
Harsh Dhamecha
|
c5991d7cd8
|
Updated C2f docstring (#3732)
Co-authored-by: Glenn Jocher <glenn.jocher@ultralytics.com>
|
2 years ago |
Adrian Boguszewski
|
e2f9bba964
|
Update OpenVINO requirement to `openvino>=2023.0` (#3722)
Co-authored-by: Glenn Jocher <glenn.jocher@ultralytics.com>
|
2 years ago |
|
Glenn Jocher
|
c55a98ab8e
|
`ultralytics 8.0.133` add `torchvision` compatibility check (#3703)
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
|
2 years ago |
|
Adrian Boguszewski
|
0821ccb618
|
Changed default device to AUTO in OpenVINO (#3699)
Co-authored-by: Glenn Jocher <glenn.jocher@ultralytics.com>
|
2 years ago |
triple Mu
|
82920ef7ec
|
Replace an easy way to construct a `ncnn.Mat` (#3638)
|
2 years ago |
|
Glenn Jocher
|
495edc261f
|
`ultralytics 8.0.132` add AutoBackend NCNN inference (#3615)
Co-authored-by: triple Mu <gpu@163.com>
|
2 years ago |
|
Colin Wong
|
23c7cd4c9f
|
Avoid Python 3.9 with_stem (#3589)
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Glenn Jocher <glenn.jocher@ultralytics.com>
|
2 years ago |
|
Glenn Jocher
|
abb0939fe8
|
`ultralytics 8.0.129` add YOLOv8 Tencent NCNN export (#3529)
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
|
2 years ago |
|
Glenn Jocher
|
8a11eda4a9
|
`ultralytics 8.0.125` NMS speed improvements (#3463)
Co-authored-by: Compunet <117437050+dme-compunet@users.noreply.github.com>
|
2 years ago |
|
Glenn Jocher
|
cd0bf05aa4
|
Cap head scaling with `min(nc, 100)` (#3436)
|
2 years ago |
Kayzwer
|
51d8cfa9c3
|
Fix RT-DETR exported onnx model (#3317)
Co-authored-by: Glenn Jocher <glenn.jocher@ultralytics.com>
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
|
2 years ago |
|
Laughing
|
9d1e5567de
|
[RTDETR]Fix val loss (#3280)
|
2 years ago |
|
Glenn Jocher
|
d8701b42ca
|
`ultralytics 8.0.120` CLI support for SAM, RTDETR (#3273)
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
|
2 years ago |
|
Laughing
|
eda7758521
|
[RTDETR]fix "Grad strides do not match bucket view strides" when training with DDP (#3255)
|
2 years ago |
|
Glenn Jocher
|
e78fb683f4
|
`default.yaml` type comments (#3237)
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
|
2 years ago |
|
Glenn Jocher
|
bd0f7ecf6f
|
Docstrings arguments cleanup (#3229)
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
|
2 years ago |
Ayush Chaurasia
|
a0ba8ef5f0
|
Add RTDETR Trainer (#2745)
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Glenn Jocher <glenn.jocher@ultralytics.com>
Co-authored-by: Laughing-q <1185102784@qq.com>
Co-authored-by: Kayzwer <68285002+Kayzwer@users.noreply.github.com>
Co-authored-by: Laughing <61612323+Laughing-q@users.noreply.github.com>
|
2 years ago |
|
Ayush Chaurasia
|
52ae65ac3b
|
Optionally `train_args` on attempt_load_weights() (#2928)
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Glenn Jocher <glenn.jocher@ultralytics.com>
|
2 years ago |
|
Laughing
|
c050b2d1a8
|
Task augment (#2924)
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
|
2 years ago |
|
Glenn Jocher
|
fd94d312da
|
`ultralytics 8.0.111` refactored `model.loss()` method (#2911)
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Snyk bot <snyk-bot@snyk.io>
|
2 years ago |
|
Ayush Chaurasia
|
facb7861cf
|
Move loss calculation to head (#2874)
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Laughing-q <1185102784@qq.com>
Co-authored-by: Glenn Jocher <glenn.jocher@ultralytics.com>
Co-authored-by: Laughing <61612323+Laughing-q@users.noreply.github.com>
|
2 years ago |
|
Ayush Chaurasia
|
7f077f7654
|
Import `ultralytics.nn.tasks` as `ultralytics.nn` (#2889)
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Glenn Jocher <glenn.jocher@ultralytics.com>
|
2 years ago |
|
Glenn Jocher
|
a2bb42dfe9
|
Fix `Conv2()` fusing (#2885)
|
2 years ago |
|
Ayush Chaurasia
|
527a97759b
|
Revert loss head PR (#2873)
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Glenn Jocher <glenn.jocher@ultralytics.com>
|
2 years ago |
|
Glenn Jocher
|
441e67d330
|
Add Conv2() module (#2820)
|
2 years ago |
|
Ayush Chaurasia
|
d19c5b6ce8
|
Move loss to task heads (#2825)
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Laughing-q <1185102784@qq.com>
|
2 years ago |
