c161557563
Default `OMP_NUM_THREADS=8` ( #6770 )
2 years ago
47f265dde2
Update min warmup iterations from 1k to 100 ( #6768 )
2 years ago
4bab56526b
Update tutorial.ipynb (2 CPUs, 12.7 GB RAM, 42.2/166.8 GB disk) ( #6767 )
2 years ago
66aaf5184d
Pre-commit table fix ( #6744 )
2 years ago
3752807c0b
YOLOv5 v6.1 release ( #6739 )
2 years ago
2692e67c5f
Updated VOC hyperparameters ( #6732 )
* Update hyps
* Update hyp.VOC.yaml
* Update pathlib
* Update hyps
* Update hyps
* Update hyps
* Update hyps
2 years ago
a936f5f219
Switch default LR scheduler from cos to linear ( #6729 )
* Switch default LR scheduler from cos to linear
Based on empirical results of training both ways on all YOLOv5 models.
* linear bug fix
2 years ago
dbbb57cf0b
GH: add PR template ( #6482 )
* GH: add PR template
* Update CONTRIBUTING.md
* Update PULL_REQUEST_TEMPLATE.md
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Update PULL_REQUEST_TEMPLATE.md
* Update PULL_REQUEST_TEMPLATE.md
* Update PULL_REQUEST_TEMPLATE.md
* Update PULL_REQUEST_TEMPLATE.md
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
c43f135571
Update `nw` to `max(nd, 1)` ( #6714 )
2 years ago
4de8b24881
Suppress `torch` AMP-CPU warnings ( #6706 )
This is a torch bug, but they seem unable or unwilling to fix it so I'm creating a suppression in YOLOv5.
Resolves https://github.com/ultralytics/yolov5/issues/6692
2 years ago
de9c25b35e
Use `export_formats()` in export.py ( #6705 )
* Use `export_formats()` in export.py
* list fix
2 years ago
a297efc383
Edge TPU inference fix ( #6686 )
* refactor: use edgetpu flag
* fix: remove bitwise and assignation to tflite
* Cleanup and fix tflite
* Cleanup
Co-authored-by: Glenn Jocher <glenn.jocher@ultralytics.com>
2 years ago
0365379016
Fix floating point in number of workers `nw` ( #6701 )
Integer division by a float yields a (rounded) float. This causes
the dataloader to crash when creating a range.
2 years ago
56697c43ff
Fix CoreML P6 inference ( #6700 )
* Fix CoreML P6 inference
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
2 years ago
3670d91b05
Fix `--evolve --bucket gs://...` ( #6698 )
2 years ago
66e5d794c7
Fix TF exports >= 2GB ( #6292 )
* Fix exporting saved_model: pb exceeds 2GB
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Replace TF v1.x API with TF v2.x API for saved_model export
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Clean up
* Remove lambda in tf.function()
* Revert "Remove lambda in tf.function()" to be compatible with TF v2.4
This reverts commit 46c7931f11
2 years ago
07221f1591
FROM nvcr.io/nvidia/pytorch:21.10-py3 ( #6677 )
Revert to 21.10 on autobuild fail
2 years ago
18b5081166
FROM nvcr.io/nvidia/pytorch:21.10-py3 ( #6673 )
Reordered installation may help reduce resource usage in autobuild
2 years ago
25456284b4
Update Dockerfile reorder installs ( #6672 )
Also `nvidia-tensorboard-plugin-dlprof`, `nvidia-tensorboard` are no longer installed in NVCR base.
2 years ago
09d263b376
FROM nvcr.io/nvidia/pytorch:21.10-py3 ( #6671 )
22.10 returns 'no space left on device' error message.
Seems like a bug at docker. Raised issue in https://github.com/docker/hub-feedback/issues/2209
2 years ago
bb054d5f55
FROM nvcr.io/nvidia/pytorch:22.01-py3 ( #6670 )
2 years ago
9a8ebe671c
Update Dockerfile `torch==1.10.2+cu113` ( #6669 )
2 years ago
2e5c67e537
Robust `scipy.cluster.vq.kmeans` too few points ( #6668 )
* Handle `scipy.cluster.vq.kmeans` too few points
Resolves #6664
* Update autoanchor.py
* Cleanup
2 years ago
7b80545e8e
Add `--cache val` ( #6663 )
New `--cache val` argument will cache validation set only into RAM. Should help multi-GPU training speeds without consuming as much RAM as full `--cache ram`.
2 years ago
1ff43702a8
Validate with 2x `--workers` single-GPU/CPU fix ( #6659 )
Fix for #6658 for single-GPU and CPU training use cases
2 years ago
6769021901
Validate with 2x `--workers` ( #6658 )
2 years ago
ee6c70ae43
Fix yolov3.yaml remove list ( #6655 )
Per https://github.com/ultralytics/yolov3/issues/1887#issuecomment-1041135181
2 years ago
ca0a00784d
Fixed wandb logger KeyError ( #6637 )
2 years ago
538930023c
Fix ConfusionMatrix scale `vmin=0.0` ( #6638 )
Fix attempt for https://github.com/ultralytics/yolov5/issues/6626
2 years ago
a45e472358
YOLOv5 Export Benchmarks ( #6613 )
* Add benchmarks.py
* Update
* Add requirements
* Updates
* Updates
* Updates
* Updates
* Updates
* Updates
* dataset autodownload from root
* Update
* Redirect to /dev/null
* sudo --help
* Cleanup
* Add exports pd df
* Updates
* Updates
* Updates
* Cleanup
* dir handling fix
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Cleanup
* Cleanup2
* Cleanup3
* Cleanup model_type
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
2 years ago
96d8f86085
W&B: don't log media in evolve ( #6617 )
2 years ago
51f7750aad
W&B: Improve resume stability ( #6611 )
* log best.pt metrics at train end
* update
* Update __init__.py
* flush callbacks when using evolve
* remember batch size on resuming
* Update train.py
* improve stability of resume
Co-authored-by: Glenn Jocher <glenn.jocher@ultralytics.com>
2 years ago
3eef80457f
Add YOLOv5n to Reproduce section ( #6619 )
2 years ago
b40bdd6b2d
YouTube dependency fix `youtube_dl==2020.12.2` ( #6612 )
Per https://github.com/ultralytics/yolov5/issues/5860#issuecomment-1035320018 by @hdnh2006
2 years ago
cb2ad9f685
Fix `ROOT / data` when running W&B `log_dataset()` ( #6606 )
* Fix missing data folder when running log_dataset
* Use ROOT/'data'
* PEP8 whitespace
2 years ago
c21da596f3
Fix `hyp_evolve.yaml` indexing bug ( #6604 )
* Fix `hyp_evolve.yaml` indexing bug
Bug caused hyp_evolve.yaml to display latest generation result rather than best generation result.
* Update plots.py
* Update general.py
* Update general.py
* Update general.py
2 years ago
a5c9057dcc
Update train.py
2 years ago
364b7c5111
Fix YouTube dislike button bug in `pafy` package ( #6603 )
Per https://github.com/ultralytics/yolov5/issues/6583#issuecomment-1034421945 by @alicera
2 years ago
9cf80b7f60
return `opt` from `train.run()` ( #6581 )
2 years ago
9c513ca629
Add `DATASETS_DIR` global in general.py ( #6578 )
2 years ago
f40854b61b
Allow custom` --evolve` project names ( #6567 )
* Update train.py
As see in #6463, modification on train in evolve process to allow custom save directory.
* fix val
* PEP8
whitespace around operator
* Cleanup
Co-authored-by: Glenn Jocher <glenn.jocher@ultralytics.com>
2 years ago
f627bc53f8
Fix `plot_labels()` colored histogram bug ( #6574 )
* Fix `plot_labels()` colored histogram bug
* Cleanup
2 years ago
22da63e30f
Fix zero-export handling with `if any(f):` ( #6569 )
* Fix zero-export handling with `if any(f):`
Partial fix for https://github.com/ultralytics/yolov5/issues/6563
* Cleanup
2 years ago
006eb40ab7
Improved AutoBatch DDP error message ( #6568 )
* Improved AutoBatch DDP error message
* Cleanup
2 years ago
b57abb17f2
Move trainloader functions to class methods ( #6559 )
* Move trainloader functions to class methods
* results = ThreadPool(NUM_THREADS).imap(self.load_image, range(n))
* Cleanup
2 years ago
dc7e093071
Edge TPU TF imports fix ( #6542 )
* Edge TPU TF imports fix
Fix for https://github.com/ultralytics/yolov5/issues/6535#issuecomment-1030631526
* Update common.py
2 years ago
cba4303d32
Fix 6 Flake8 issues ( #6541 )
* F541
* F821
* F841
* E741
* E302
* E722
* Apply suggestions from code review
* Update general.py
* Update datasets.py
* Update export.py
* Update plots.py
* Update plots.py
Co-authored-by: Glenn Jocher <glenn.jocher@ultralytics.com>
2 years ago
e1a6a0b100
flake8: code meanings ( #6481 )
2 years ago
aff0281969
Load checkpoint on CPU instead of on GPU ( #6516 )
* Load checkpoint on CPU instead of on GPU
* refactor: simplify code
* Cleanup
* Update train.py
Co-authored-by: Glenn Jocher <glenn.jocher@ultralytics.com>
2 years ago
8fcdf3b60b
Fixing minor multi-streaming issues with TensoRT engine ( #6504 )
* Update batch-size in model.warmup() + indentation for logging inference results
* These changes are in response to PR comments
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
2 years ago
Glenn Jocher
Jirka Borovec
Raffaele Galliera
Samuel Yvon
Jiacong Fang
imyhxy
Ayush Chaurasia
or-toledano
Christoph Fürbacher
Matthias Vadcard
bilzard
greg2451