Commit Graph

  • a19406b39d
    Update minimum Python>=3.7.0 (#7247) Glenn Jocher 2022-04-02 15:05:00 +0200
  • 37675e110f
    Fix `www.youtube.com` hostname (#7242) Glenn Jocher 2022-04-01 21:38:49 +0200
  • 71621df875
    Create CODE_OF_CONDUCT.md (#7233) Glenn Jocher 2022-04-01 00:24:37 +0200
  • 734ab033fd
    SavedModel TF Serve Fix (#7228) Glenn Jocher 2022-04-01 00:07:23 +0200
  • 4d157f578a
    Update .pre-commit-config.yaml (#7230) Glenn Jocher 2022-03-31 17:26:34 +0200
  • 2c3221844b
    CLI `fire` prep updates (#7229) Glenn Jocher 2022-03-31 17:11:43 +0200
  • c3d5ac151e
    precommit: yapf (#5494) Jirka Borovec 2022-03-31 23:52:34 +0900
  • df9008ee69
    Add train.py `--name cfg` option (#7202) Glenn Jocher 2022-03-31 13:17:22 +0200
  • c94736acec
    `ENV OMP_NUM_THREADS=8` (#7215) Glenn Jocher 2022-03-30 16:01:55 +0200
  • 9c1e91aea2
    Update tutorial.ipynb (#7212) Glenn Jocher 2022-03-30 12:53:49 +0200
  • cf4f3c3455
    yolo.py profiling updates (#7178) Glenn Jocher 2022-03-29 10:15:53 +0200
  • d51f9b2ff6
    Bump actions/cache from 2.1.7 to 3 (#7175) dependabot[bot] 2022-03-28 10:42:19 +0200
  • ee77632393
    Revert `C3()` change (#7172) Glenn Jocher 2022-03-28 02:31:00 +0200
  • 1832264dd6 Update Glenn Jocher 2022-03-28 02:26:44 +0200
  • b2194b9015
    `yolo.py --profile` updates (#7170) Glenn Jocher 2022-03-27 20:24:42 +0200
  • 7830e91b9a
    `yolo.py --profile` default GPU batch size 16 Glenn Jocher 2022-03-26 16:53:42 +0100
  • 3373aab56c
    NMS unused variable fix (#7161) Glenn Jocher 2022-03-26 16:52:58 +0100
  • e19f87eb4b
    Sidestep `os.path.relpath()` Windows bug (#7158) Glenn Jocher 2022-03-26 14:18:53 +0100
  • 26bfd44465
    Adjust NMS time limit warning to batch size (#7156) Glenn Jocher 2022-03-26 11:45:28 +0100
  • 7a2a11893b
    Add Architecture Summary to README Tutorials (#7146) Glenn Jocher 2022-03-25 14:45:23 +0100
  • a4c661873f
    Fix `detect.py --view-img` for non-ASCII paths (#7093) Zengyf-CVer 2022-03-25 20:40:55 +0800
  • d115bbf509
    Fix `cv2.imwrite` on non-ASCII paths (#7139) RcINS 2022-03-25 20:25:30 +0800
  • a0a4adf6de
    Add PyTorch Hub `results.save(labels=False)` option (#7129) Glenn Jocher 2022-03-24 11:31:22 +0100
  • bc3ed957ce
    `np.fromfile()` Chinese image paths fix (#6979) yeshanliu 2022-03-23 22:35:15 +0800
  • c3ae4e4af6
    Multi-threaded VisDrone and VOC downloads (#7108) Glenn Jocher 2022-03-23 01:19:37 +0100
  • ecc2c7ba73
    Remove named arguments where possible (#7105) Glenn Jocher 2022-03-22 20:44:07 +0100
  • 6134ec5d94
    Model summary `pathlib` fix (#7104) Glenn Jocher 2022-03-22 20:05:07 +0100
  • ee0b3b2a95
    Update benchmarks significant digits (#7103) Glenn Jocher 2022-03-22 18:02:35 +0100
  • 05aae17333
    `torch.split()` 1.7.0 compatibility fix (#7102) Glenn Jocher 2022-03-22 17:36:05 +0100
  • a600baed8e
    Update model summary to display model name (#7101) Glenn Jocher 2022-03-22 15:41:19 +0100
  • a2d617ece9
    Update loss for FP16 `tobj` (#7088) Glenn Jocher 2022-03-21 19:18:34 +0100
  • 6f128031d0
    Update loss.py with `if self.gr < 1:` (#7087) Glenn Jocher 2022-03-21 18:35:36 +0100
  • 9b771a3e71
    Revert "Update detect.py non-inplace with `y.tensor_split()` (#7062)" (#7074) Glenn Jocher 2022-03-21 09:33:39 +0100
  • 9cd89b75cc
    Fix2 `check_anchor_order()` in pixel-space not grid-space (#7067) Glenn Jocher 2022-03-20 18:55:13 +0100
  • 178c109576
    Add non-zero `da` `check_anchor_order()` condition (#7066) Glenn Jocher 2022-03-20 18:37:27 +0100
  • 9e75cbf4c1
    Update __init__.py (#7065) Glenn Jocher 2022-03-20 18:17:04 +0100
  • e278fd63ec
    Update W&B message to `LOGGER.info()` (#7064) Glenn Jocher 2022-03-20 18:15:22 +0100
  • 0529b77232
    Update common.py lists for tuples (#7063) Glenn Jocher 2022-03-20 18:03:37 +0100
  • d5e363f29d
    Update detect.py non-inplace with `y.tensor_split()` (#7062) Glenn Jocher 2022-03-20 18:02:05 +0100
  • f327eee614
    Fix `check_anchor_order()` in pixel-space not grid-space (#7060) Glenn Jocher 2022-03-20 16:27:51 +0100
  • 529fbc1814
    Use PIL to eliminate chroma subsampling in crops (#7008) Philip Gutjahr 2022-03-20 15:46:29 +0100
  • 9f4d11379b
    Fix incomplete URL substring sanitation (#7056) Glenn Jocher 2022-03-20 14:18:05 +0100
  • 6843ea5d7f
    Create SECURITY.md (#7054) Glenn Jocher 2022-03-20 13:55:32 +0100
  • 9ebec7885f
    Update Dockerfile to `git clone` instead of `COPY` (#7053) Glenn Jocher 2022-03-20 13:51:22 +0100
  • b0ba101ac0
    `ComputeLoss()` indexing/speed improvements (#7048) Glenn Jocher 2022-03-20 01:04:48 +0100
  • 4effd064b1
    Consistent saved_model output format (#7032) Mrinal Jain 2022-03-18 07:29:24 -0400
  • 7c6a33564a
    fix: add default PIL font as fallback (#7010) Max Strobel 2022-03-17 16:37:09 +0100
  • 3f634d43c8
    Conditional `Timeout()` by OS (disable on Windows) (#7013) Glenn Jocher 2022-03-16 15:33:54 +0100
  • c09fb2aa95
    Update TQDM bar format (#6988) Glenn Jocher 2022-03-15 16:32:56 +0100
  • 932dc78496
    YOLOv5 Export Benchmarks for GPU (#6963) Glenn Jocher 2022-03-14 15:07:13 +0100
  • 99de551f97
    pt model to cpu on TF export Glenn Jocher 2022-03-14 12:41:06 +0100
  • 2d45de617e
    Model `ema` key backward compatibility fix (#6972) Glenn Jocher 2022-03-14 10:54:51 +0100
  • c13d4ce7ef
    EdgeTPU optimizations (#6808) paradigm 2022-03-12 16:15:09 +0100
  • 701e1177ac
    Tensor initialization on device improvements (#6959) Glenn Jocher 2022-03-12 14:00:48 +0100
  • 52c1399fdc
    DetectMultiBackend() return `device` update (#6958) Glenn Jocher 2022-03-12 13:16:29 +0100
  • c84dd27d62
    New val.py `cuda` variable (#6957) Glenn Jocher 2022-03-12 12:57:08 +0100
  • c6b4f84fd1
    Update Dockerfile `torch==1.11.0+cu113` (#6954) Glenn Jocher 2022-03-12 00:45:07 +0100
  • b94b59e199
    DetectMultiBackend() `--half` handling (#6945) Glenn Jocher 2022-03-11 16:31:52 +0100
  • 84efa62b2d
    Fix PyTorch Hub export inference shapes (#6949) Glenn Jocher 2022-03-11 16:18:40 +0100
  • caf7ad0500
    Allow 3-point segments (#6938) Glenn Jocher 2022-03-10 18:41:47 +0100
  • 055e72af5b
    Optimize PyTorch 1.11.0 compatibility update (#6933) Glenn Jocher 2022-03-10 12:58:41 +0100
  • d3d9cbce22
    PyTorch 1.11.0 compatibility updates (#6932) Glenn Jocher 2022-03-10 12:41:06 +0100
  • 6dd82c0252
    Move `git_describe()` to general.py (#6918) Glenn Jocher 2022-03-09 18:22:53 +0100
  • e6e36aac10
    Update bytes to GB with bitshift (#6886) Glenn Jocher 2022-03-07 19:26:37 +0100
  • acc58c1dcf
    Fix TRT `max_workspace_size` deprecation notice (#6856) Glenn Jocher 2022-03-07 13:52:53 +0100
  • a5a1760ea6
    Bump actions/checkout from 2 to 3 (#6881) dependabot[bot] 2022-03-07 13:49:27 +0100
  • c8a589920e
    Bump actions/setup-python from 2 to 3 (#6880) dependabot[bot] 2022-03-07 13:48:59 +0100
  • 596de6d5a0
    Default FP16 TensorRT export (#6798) DavidB 2022-03-07 03:21:16 +0700
  • 7e98b4801a
    Update yolov5s.yaml (#6865) vnekat 2022-03-07 00:50:01 +0530
  • 4728840745
    Update `--cache disk` deprecate `*_npy/` dirs (#6876) Glenn Jocher 2022-03-06 16:16:17 +0100
  • 8a66ebad44
    Add `*.ts` to `VID_FORMATS` (#6859) Glenn Jocher 2022-03-04 14:10:13 +0100
  • 601dbb83f0
    AutoAnchor improved initialization robustness (#6854) Glenn Jocher 2022-03-04 10:32:18 +0100
  • bcc92e2169
    Update sweep.yaml (#6825) Louis Combaldieu 2022-03-04 09:39:23 +0100
  • 63ddb6f0d0
    Update autoanchor.py (#6794) Glenn Jocher 2022-02-26 19:15:12 +0100
  • c2403eb04c
    Update PULL_REQUEST_TEMPLATE.md (#6783) Glenn Jocher 2022-02-25 14:01:58 +0100
  • 9ec51a6396
    YOLOv5s6 params FLOPs fix (#6782) Glenn Jocher 2022-02-25 13:59:21 +0100
  • 0f819919ad
    Update EMA decay `tau` (#6769) Glenn Jocher 2022-02-25 12:33:09 +0100
  • b2adc7c39a
    Fix export for 1-channel images (#6780) Louis Combaldieu 2022-02-25 10:56:37 +0100
  • cea994b3f6
    Update hyp.VOC.yaml (#6772) Glenn Jocher 2022-02-25 00:14:17 +0100
  • 741cd0eb23
    Update tutorial.ipynb (#6771) Glenn Jocher 2022-02-25 00:12:27 +0100
  • c161557563
    Default `OMP_NUM_THREADS=8` (#6770) Glenn Jocher 2022-02-25 00:09:14 +0100
  • 47f265dde2
    Update min warmup iterations from 1k to 100 (#6768) Glenn Jocher 2022-02-24 23:37:25 +0100
  • 4bab56526b
    Update tutorial.ipynb (2 CPUs, 12.7 GB RAM, 42.2/166.8 GB disk) (#6767) Glenn Jocher 2022-02-24 18:29:59 +0100
  • 66aaf5184d
    Pre-commit table fix (#6744) Glenn Jocher 2022-02-22 19:35:31 +0100
  • 3752807c0b
    YOLOv5 v6.1 release (#6739) Glenn Jocher 2022-02-22 12:35:24 +0100
  • 2692e67c5f
    Updated VOC hyperparameters (#6732) Glenn Jocher 2022-02-22 12:11:18 +0100
  • a936f5f219
    Switch default LR scheduler from cos to linear (#6729) Glenn Jocher 2022-02-21 14:23:27 +0100
  • dbbb57cf0b
    GH: add PR template (#6482) Jirka Borovec 2022-02-21 12:36:43 +0100
  • c43f135571
    Update `nw` to `max(nd, 1)` (#6714) Glenn Jocher 2022-02-20 21:47:46 +0100
  • 4de8b24881
    Suppress `torch` AMP-CPU warnings (#6706) Glenn Jocher 2022-02-19 16:48:33 +0100
  • de9c25b35e
    Use `export_formats()` in export.py (#6705) Glenn Jocher 2022-02-19 16:08:33 +0100
  • a297efc383
    Edge TPU inference fix (#6686) Raffaele Galliera 2022-02-19 08:10:07 -0600
  • 0365379016
    Fix floating point in number of workers `nw` (#6701) Samuel Yvon 2022-02-18 15:06:21 -0500
  • 56697c43ff
    Fix CoreML P6 inference (#6700) Glenn Jocher 2022-02-18 20:38:33 +0100
  • 3670d91b05
    Fix `--evolve --bucket gs://...` (#6698) Glenn Jocher 2022-02-18 19:15:37 +0100
  • 66e5d794c7
    Fix TF exports >= 2GB (#6292) Jiacong Fang 2022-02-18 23:36:45 +0800
  • 07221f1591
    FROM nvcr.io/nvidia/pytorch:21.10-py3 (#6677) Glenn Jocher 2022-02-17 16:05:30 +0100
  • 18b5081166
    FROM nvcr.io/nvidia/pytorch:21.10-py3 (#6673) Glenn Jocher 2022-02-17 15:42:47 +0100
  • 25456284b4
    Update Dockerfile reorder installs (#6672) Glenn Jocher 2022-02-17 15:10:19 +0100
  • 09d263b376
    FROM nvcr.io/nvidia/pytorch:21.10-py3 (#6671) Glenn Jocher 2022-02-17 14:51:53 +0100