Commit Graph

  • affa284352
    Refactor/reduce G/C/D/IoU `if: else` statements (#6087) Chen Gen 2021-12-31 05:59:29 +0800
  • 7b6938d5b5
    Log best results (#6085) Awsaf 2021-12-31 03:47:53 +0600
  • 9155eb8641
    W&B: Log best results after training ends (#6120) Ayush Chaurasia 2021-12-31 01:13:19 +0530
  • db6ec66a60
    W&B: track batch size after autobatch (#6039) Ayush Chaurasia 2021-12-23 05:23:50 -0800
  • c72270c076
    Init tensor directly on device (#6068) Deep Patel 2021-12-23 07:49:00 -0500
  • afa5cfb0f8
    Reduce G/D/CIoU logic operations (#6074) JieLi 2021-12-23 18:53:00 +0800
  • 95c7bc25d3
    OpenVINO Export (#6057) Glenn Jocher 2021-12-22 20:29:48 +0100
  • dc54ed5763
    `--freeze` fix (#6044) Glenn Jocher 2021-12-20 18:24:07 +0100
  • b8a4babd60
    Simplify `set_logging()` indexing (#6042) Glenn Jocher 2021-12-20 17:42:52 +0100
  • 0db9d5b6a2
    Kaggle `LOGGER` fix (#6041) Glenn Jocher 2021-12-20 17:30:46 +0100
  • 26f0415287
    Add dataset source citations (#6032) Glenn Jocher 2021-12-19 15:19:04 +0100
  • abbdd4802e
    train -> val comment fix (#6024) Glenn Jocher 2021-12-17 16:43:00 +0100
  • 361705d9be
    Multi-layer capable `--freeze` argument (#6019) Felix You 2021-12-17 22:42:26 +0800
  • 407a905747
    Check TensorRT>=8.0.0 version (#6021) Glenn Jocher 2021-12-17 14:59:46 +0100
  • c1249a47c7
    *.torchscript inference `self.jit` fix (#6007) Glenn Jocher 2021-12-16 14:10:54 +0100
  • 628817dfae
    Recommend `jar xf file.zip` for zips (#5993) Glenn Jocher 2021-12-15 17:19:19 +0100
  • da9a1b719b
    Allow `--weights URL` (#5991) Glenn Jocher 2021-12-15 15:27:08 +0100
  • b7d18f3636
    Increase `ar_thr` from 20 to 100 for better detection on slender (high aspect ratio) objects (#5556) Mrinal Jain 2021-12-15 19:12:23 +0530
  • c9a46a60e0
    Update callbacks.py with `__init__()` (#5979) Glenn Jocher 2021-12-14 15:47:49 +0100
  • d699c21c75
    YOLOv5s6 params and FLOPs fix (#5977) Glenn Jocher 2021-12-14 11:24:39 +0100
  • 2d0c6afbfe
    make parameter ignore epochs (#5972) jinmc 2021-12-14 19:18:34 +0900
  • e8ef8fb1ca
    `pretrained=False` fix (#5966) Glenn Jocher 2021-12-13 13:32:27 +0100
  • 19c56e60b1
    Fix `imgsz` bug (#5948) Diego Montes 2021-12-12 17:39:14 -0500
  • 8f354362cd
    Fix Detections class `tolist()` method (#5945) Yono Mittlefehldt 2021-12-11 18:40:37 +0100
  • 8f875d93a2
    Refactor NUM_THREADS (#5954) Glenn Jocher 2021-12-11 14:32:18 +0100
  • 2c6317547a
    Add nms and agnostic nms to export.py (#5938) Diego Montes 2021-12-10 12:24:32 -0500
  • a42af30d8a
    Update `strip_optimizer()` (#5949) iumyx2612 2021-12-11 00:06:27 +0700
  • 922fbd8209
    fix .gitignore not tracking existing folders (#5946) Pascal Maillard 2021-12-10 17:32:09 +0100
  • c45f9f678d
    Make `select_device()` robust to `batch_size=-1` (#5940) Felix You 2021-12-10 22:27:20 +0800
  • 4fb6dd4b26
    Fix ONNX opset inconsistency with parseargs and run args (#5937) Diego Montes 2021-12-09 17:10:16 -0500
  • 5bdb28ed10
    Default PyTorch Hub to `autocast(False)` (#5926) Glenn Jocher 2021-12-08 23:15:14 +0100
  • c77a5a84e3
    Absolute '/content/sample_data' (#5922) Glenn Jocher 2021-12-08 17:31:31 +0100
  • a3d5f1d3e3
    Revert "Update `plot_lr_scheduler()` (#5864)" (#5920) Glenn Jocher 2021-12-08 16:46:24 +0100
  • 7d56d45124
    Add hardware checks to `notebook_init()` (#5919) Glenn Jocher 2021-12-08 14:57:03 +0100
  • 581dc301a7
    Add ONNX inference providers (#5918) Glenn Jocher 2021-12-08 13:37:33 +0100
  • 554f782537
    Add *.engine (TensorRT extensions) to .gitignore (#5911) greg2451 2021-12-07 18:20:16 +0100
  • 3f152e5807
    Update wandb_utils.py (#5908) Ayush Chaurasia 2021-12-07 20:39:11 +0530
  • fa05f8c977
    `Detections().tolist()` explicit argument fix (#5907) Li Zeng 2021-12-07 16:01:41 +0100
  • 1075488d89
    Single-command multiple-model export (#5882) Glenn Jocher 2021-12-04 16:28:40 +0100
  • 7bf04d9bbf
    `AutoShape()` models as `DetectMultiBackend()` instances (#5845) Glenn Jocher 2021-12-04 15:00:07 +0100
  • d885799c71
    Update `nl` after `cutout()` (#5873) Glenn Jocher 2021-12-03 15:28:14 +0100
  • 360eec6910
    Update `plot_lr_scheduler()` (#5864) Can 2021-12-03 20:37:45 +0800
  • 92a7391039
    Add `--workers 8` argument to val.py (#5857) iumyx2612 2021-12-02 22:49:50 +0700
  • 00e308f7be
    Update TorchScript suffix to `*.torchscript` (#5856) Glenn Jocher 2021-12-02 16:06:45 +0100
  • 30db14fea8
    Update val.py (#5838) Vishnu Pradeep 2021-12-02 13:20:58 +0100
  • 1679aacdc7
    Update `LoadImages` `ret_val=False` handling (#5852) gggmt 2021-12-02 17:57:39 +0800
  • e8f8f2b903
    Avoid inplace modifying`imgs` in `LoadStreams` (#5850) Yu Zhang 2021-12-02 17:51:19 +0800
  • bc484579d7
    Handle non-TTY `wandb.errors.UsageError` (#5839) Glenn Jocher 2021-12-01 15:38:02 +0100
  • a4207a202d
    Fix TensorRT potential unordered binding addresses (#5826) imyhxy 2021-11-30 20:52:22 +0800
  • 5ca5dd4c87
    Update `dataset_stats()` to `cv2.INTER_AREA` (#5821) Glenn Jocher 2021-11-29 13:15:36 +0100
  • 8277033b65
    Bump actions/cache from 2.1.6 to 2.1.7 (#5816) dependabot[bot] 2021-11-29 12:04:36 +0100
  • 94d8fec6d8
    GCP VM from Image example (#5814) Glenn Jocher 2021-11-28 18:12:46 +0100
  • fcd180d336
    Refactor new `model.warmup()` method (#5810) Glenn Jocher 2021-11-27 12:29:45 +0100
  • 7c6bae0ae6
    Remove NCOLS from tqdm (#5804) Glenn Jocher 2021-11-26 13:37:28 +0100
  • 53349dac8e
    Scope TF imports in `DetectMultiBackend()` (#5792) Phil2020 2021-11-25 16:54:00 +0000
  • f2ca30a407
    W&B: refactor W&B tables (#5737) Ayush Chaurasia 2021-11-25 21:49:21 +0530
  • 4c7b2bdc30
    TorchScript `torch==1.7.0` Path support (#5781) rockstarr 2021-11-25 18:18:30 +0800
  • c55e674ac3
    fix the path error in export.py (#5778) rockstarr 2021-11-25 15:42:15 +0800
  • 4ca4aec46f
    Do not save hyp.yaml and opt.yaml on evolve (#5775) Glenn Jocher 2021-11-24 23:25:51 +0100
  • 7a39803476
    Export, detect and validation with TensorRT engine file (#5699) imyhxy 2021-11-22 21:58:07 +0800
  • f17c86b7f0
    Save *.npy features on detect.py `--visualize` (#5701) Zengyf-CVer 2021-11-22 03:21:44 +0800
  • d6ae1c835a Created using Colaboratory Glenn Jocher 2021-11-21 19:26:07 +0100
  • 5185981993
    SECURITY.md (#5695) IL2006 2021-11-20 08:26:48 +0800
  • 46daa7b78d
    Remove `.autoshape()` method (#5694) Glenn Jocher 2021-11-20 01:11:36 +0100
  • 36d12a500e
    Explicitly compute TP, FP in val.py (#5727) Glenn Jocher 2021-11-20 01:04:56 +0100
  • eb51ffdcac
    Prune unused imports (#5711) Jirka Borovec 2021-11-19 13:32:53 +0100
  • 8df64a9122
    `NUM_THREADS` leave at least 1 CPU free (#5706) Glenn Jocher 2021-11-18 14:53:42 +0100
  • 562191f575
    Update export.py docstring (#5689) Glenn Jocher 2021-11-17 15:18:50 +0100
  • 47fac9ff73
    Update train.py comment to 'Model attributes' (#5670) Glenn Jocher 2021-11-16 13:58:15 +0100
  • 0453b758e7
    TFDetect dynamic anchor count assignment fix (#5668) Nrupatunga 2021-11-16 17:06:00 +0530
  • e80a09bbfa
    Update ci-testing.yml to Python 3.9 (#5660) Glenn Jocher 2021-11-15 20:15:50 +0100
  • fb19561f98
    Bump pip from 19.2 to 21.1 in /utils/google_app_engine (#5661) dependabot[bot] 2021-11-15 20:10:29 +0100
  • c2523be634
    Replace 2 `transpose()` with 1 `permute` in TransformerBlock()` (#5645) Ding Yiwei 2021-11-15 17:06:18 +0800
  • 540ef0dd30
    W&B refactor, handle exceptions, CI example (#5618) Ayush Chaurasia 2021-11-14 17:56:53 +0530
  • 80cfaf40ef
    AutoAnchor and AutoBatch `LOGGER` (#5635) Glenn Jocher 2021-11-13 15:40:18 +0100
  • 09d170381c
    Default DataLoader `shuffle=True` for training (#5623) Werner Duvaud 2021-11-13 12:07:32 +0000
  • 7473f0f95d
    DDP `WORLD_SIZE`-safe dataloader workers (#5631) Glenn Jocher 2021-11-12 14:48:55 +0100
  • d5b21b1ecb
    HUB dataset previews to JPEG (#5627) Glenn Jocher 2021-11-12 12:05:25 +0100
  • def7a0fd19
    Improve plots.py robustness (#5616) Glenn Jocher 2021-11-11 12:56:38 +0100
  • 69032519bc
    `np.unique()` sort fix for segments (#5609) Glenn Jocher 2021-11-11 00:15:17 +0100
  • 30bc089cbb
    Update val.py `speed` and `study` tasks (#5608) Glenn Jocher 2021-11-10 16:48:38 +0100
  • 61c50199a2
    Update train, val `tqdm` to fixed width (#5367) Glenn Jocher 2021-11-10 15:47:38 +0100
  • 27bf4282d3
    Fix `check_requirements()` resource warning allocation open file (#5602) Ayman Saleh 2021-11-10 06:51:30 -0500
  • 7ebb5e5da6
    Add `notebook_init()` to utils/__init__.py (#5488) Glenn Jocher 2021-11-09 23:03:19 +0100
  • 7207fe95e5
    FROM nvcr.io/nvidia/pytorch:21.10-py3 (#5592) Glenn Jocher 2021-11-09 17:55:57 +0100
  • 3883261143
    New `DetectMultiBackend()` class (#5549) Glenn Jocher 2021-11-09 16:45:02 +0100
  • 79bca2bf64
    `LOGGER` consolidation (#5569) Glenn Jocher 2021-11-08 16:32:15 +0100
  • 0de4a9c35d
    Add `--conf-thres` >> 0.001 warning (#5567) Glenn Jocher 2021-11-08 16:04:31 +0100
  • b8f979bafa
    Inside Ultralytics video https://youtu.be/Zgi9g1ksQHc (#5546) Glenn Jocher 2021-11-06 20:34:54 +0100
  • 3f64ad1760
    Fix `save_one_box()` (#5545) Glenn Jocher 2021-11-06 19:28:03 +0100
  • 60c8a4f696
    Fix for *.yaml emojis on load (#5543) Glenn Jocher 2021-11-06 16:03:14 +0100
  • e189fa15ea
    `intersect_dicts()` in hubconf.py fix (#5542) Glenn Jocher 2021-11-06 15:41:17 +0100
  • fa2344cdd8
    Update `models/hub/*.yaml` files for v6.0n release (#5540) Glenn Jocher 2021-11-06 15:07:45 +0100
  • 76d90d899a
    Update Issue Templates with 💡 ProTip! (#5539) Glenn Jocher 2021-11-06 13:58:12 +0100
  • cb18cac33d
    Update autobatch.py (#5538) Glenn Jocher 2021-11-06 13:49:00 +0100
  • 60e42e16c2
    Update autobatch.py (#5536) Glenn Jocher 2021-11-06 12:21:17 +0100
  • 336437998f
    Suppress ONNX export trace warning (#5437) Deep Patel 2021-11-05 14:31:53 -0400
  • d895a7f70d
    Update train.py (#5451) Wonbeom Jang 2021-11-06 03:28:53 +0900
  • 98a3fd7e8f
    Update export.py (#5471) nanmi 2021-11-06 02:26:45 +0800
  • 32b8738735
    Update `check_file()` avoid repeat URL downloads (#5526) Glenn Jocher 2021-11-05 19:22:47 +0100