Commit Graph

  • bb054d5f55
    FROM nvcr.io/nvidia/pytorch:22.01-py3 (#6670) Glenn Jocher 2022-02-17 14:20:42 +0100
  • 9a8ebe671c
    Update Dockerfile `torch==1.10.2+cu113` (#6669) Glenn Jocher 2022-02-17 13:24:44 +0100
  • 2e5c67e537
    Robust `scipy.cluster.vq.kmeans` too few points (#6668) Glenn Jocher 2022-02-17 12:55:03 +0100
  • 7b80545e8e
    Add `--cache val` (#6663) Glenn Jocher 2022-02-17 00:47:42 +0100
  • 1ff43702a8
    Validate with 2x `--workers` single-GPU/CPU fix (#6659) Glenn Jocher 2022-02-16 14:50:37 +0100
  • 6769021901
    Validate with 2x `--workers` (#6658) Glenn Jocher 2022-02-16 14:44:44 +0100
  • ee6c70ae43
    Fix yolov3.yaml remove list (#6655) Glenn Jocher 2022-02-16 10:13:05 +0100
  • ca0a00784d
    Fixed wandb logger KeyError (#6637) imyhxy 2022-02-15 00:08:36 +0800
  • 538930023c
    Fix ConfusionMatrix scale `vmin=0.0` (#6638) Glenn Jocher 2022-02-14 14:54:55 +0100
  • a45e472358
    YOLOv5 Export Benchmarks (#6613) Glenn Jocher 2022-02-12 16:05:43 +0100
  • 96d8f86085
    W&B: don't log media in evolve (#6617) Ayush Chaurasia 2022-02-12 17:34:50 +0530
  • 51f7750aad
    W&B: Improve resume stability (#6611) Ayush Chaurasia 2022-02-12 17:33:09 +0530
  • 3eef80457f
    Add YOLOv5n to Reproduce section (#6619) Glenn Jocher 2022-02-11 21:48:01 +0100
  • b40bdd6b2d
    YouTube dependency fix `youtube_dl==2020.12.2` (#6612) Glenn Jocher 2022-02-11 13:46:20 +0100
  • cb2ad9f685
    Fix `ROOT / data` when running W&B `log_dataset()` (#6606) or-toledano 2022-02-10 17:17:47 +0200
  • c21da596f3
    Fix `hyp_evolve.yaml` indexing bug (#6604) Glenn Jocher 2022-02-10 14:51:29 +0100
  • a5c9057dcc
    Update train.py Glenn Jocher 2022-02-10 14:15:18 +0100
  • 364b7c5111
    Fix YouTube dislike button bug in `pafy` package (#6603) Glenn Jocher 2022-02-10 13:13:44 +0100
  • 9cf80b7f60
    return `opt` from `train.run()` (#6581) Christoph Fürbacher 2022-02-09 15:11:52 +0100
  • 9c513ca629
    Add `DATASETS_DIR` global in general.py (#6578) Glenn Jocher 2022-02-08 22:20:44 +0100
  • f40854b61b
    Allow custom` --evolve` project names (#6567) Matthias Vadcard 2022-02-08 16:13:45 +0100
  • f627bc53f8
    Fix `plot_labels()` colored histogram bug (#6574) Glenn Jocher 2022-02-08 14:03:50 +0100
  • 22da63e30f
    Fix zero-export handling with `if any(f):` (#6569) Glenn Jocher 2022-02-08 12:20:39 +0100
  • 006eb40ab7
    Improved AutoBatch DDP error message (#6568) Glenn Jocher 2022-02-08 11:45:30 +0100
  • b57abb17f2
    Move trainloader functions to class methods (#6559) Glenn Jocher 2022-02-07 18:59:26 +0100
  • dc7e093071
    Edge TPU TF imports fix (#6542) Glenn Jocher 2022-02-05 15:45:44 +0100
  • cba4303d32
    Fix 6 Flake8 issues (#6541) Jirka Borovec 2022-02-05 15:22:59 +0100
  • e1a6a0b100
    flake8: code meanings (#6481) Jirka Borovec 2022-02-04 19:54:34 +0100
  • aff0281969
    Load checkpoint on CPU instead of on GPU (#6516) bilzard 2022-02-05 02:58:18 +0900
  • 8fcdf3b60b
    Fixing minor multi-streaming issues with TensoRT engine (#6504) greg2451 2022-02-04 18:19:37 +0100
  • 079b36d72b
    Edge TPU `tf.lite.experimental.load_delegate` fix (#6536) Glenn Jocher 2022-02-04 09:30:01 +0100
  • 9bc72a3ac2
    Edge TPU export 'list index out of range' fix (#6533) Glenn Jocher 2022-02-04 00:04:06 +0100
  • 8d05716b66
    Edge TPU compiler `sudo` fix (#6531) Glenn Jocher 2022-02-03 20:40:09 +0100
  • a82292ec53
    Social icons after text (#6473) Glenn Jocher 2022-02-03 19:55:19 +0100
  • c3e599cfda
    Update val_batch*.jpg for Chinese fonts (#6526) Glenn Jocher 2022-02-03 19:09:24 +0100
  • cb40c9afda
    TODO issues exempt from stale action (#6530) Glenn Jocher 2022-02-03 18:11:28 +0100
  • 19e0208fc9
    Update hyp.scratch-high.yaml (#6525) Glenn Jocher 2022-02-03 12:15:13 +0100
  • b73c62ebc5
    W&B: Remember batchsize on resuming (#6512) Ayush Chaurasia 2022-02-03 15:59:52 +0530
  • 4c40933266
    Suppress export.run() TracerWarnings (#6499) Glenn Jocher 2022-02-01 23:52:50 +0100
  • 842d049e1b
    Suppress `torch.jit.TracerWarning` on export (#6498) Glenn Jocher 2022-02-01 22:59:26 +0100
  • 77977e0791
    CoreML inference fix `list()` -> `sorted()` (#6496) Glenn Jocher 2022-02-01 22:34:15 +0100
  • 5e4ff195b2
    Improved `export.py` usage examples (#6495) Glenn Jocher 2022-02-01 22:06:29 +0100
  • b884ea36c4
    Simplify TF normalized to pixels (#6494) Glenn Jocher 2022-02-01 21:17:56 +0100
  • 6445a8137e
    Resolve dataset paths (#6489) Glenn Jocher 2022-02-01 15:54:51 +0100
  • 7539cd75c3
    Add Product Hunt social media icon (#6464) Glenn Jocher 2022-01-28 20:23:17 +0100
  • d8b5beb0b0
    Fix2 `select_device()` for Multi-GPU (#6461) Glenn Jocher 2022-01-28 08:18:01 +0100
  • 856d4e5733
    Fix `select_device()` for Multi-GPU (#6434) Glenn Jocher 2022-01-26 23:47:00 +0100
  • fe7de6a82d
    Revert "Remove `dataset_stats()` autodownload capability (#6303)" (#6442) Glenn Jocher 2022-01-26 19:10:59 +0100
  • 3b7ac28ed1
    Add `*.asf` video support (#6436) toschi23 2022-01-26 14:26:02 +0100
  • d5966c93f1
    Namespace `VERBOSE` env variable to `YOLOv5_VERBOSE` (#6428) Glenn Jocher 2022-01-25 14:57:27 -0800
  • ff8646cdea
    Update workflows (#6427) Glenn Jocher 2022-01-25 14:33:22 -0800
  • 16563ac5b5
    Prefer `tflite_runtime` for TFLite inference if installed (#6406) Motoki Kimura 2022-01-26 00:24:24 +0900
  • ed9bac8392
    Rename logger from 'utils.logger' to 'yolov5' (#6421) Jonathan Samelson 2022-01-25 16:21:06 +0100
  • cfecd903a3
    Update `greetings.yaml` email address (#6412) Glenn Jocher 2022-01-24 15:28:52 -0800
  • 482af479c0
    Add `detect.py` GIF video inference (#6410) Glenn Jocher 2022-01-24 13:11:11 -0800
  • 8efe97719c
    Add `stop_training=False` flag to callbacks (#6365) Matthias 2022-01-23 03:37:21 +0100
  • c43439aa31
    Add `albumentations` to Dockerfile (#6392) Glenn Jocher 2022-01-21 21:06:02 -1000
  • bd815d48df
    FROM nvcr.io/nvidia/pytorch:21.10-py3 (#6379) Glenn Jocher 2022-01-20 19:15:04 -1000
  • 1b41a1d059
    FROM nvcr.io/nvidia/pytorch:21.12-py3 (#6377) Glenn Jocher 2022-01-20 17:59:41 -1000
  • 9bcc32a5bf
    Flush callbacks when on `--evolve` (#6374) Ayush Chaurasia 2022-01-21 04:52:47 +0530
  • e1893c894a
    `DEVICE_COUNT` instead of `WORLD_SIZE` to calculate `nw` (#6324) sitecao 2022-01-20 18:06:26 -0500
  • 4e841b9b16
    Reuse `de_parallel()` rather than `is_parallel()` (#6354) imyhxy 2022-01-21 04:50:17 +0800
  • 9708cf56ea
    New environment variable `VERBOSE` (#6353) johnk2hawaii 2022-01-19 14:32:19 -1000
  • 750c42e43e
    `export.py` automatic `forward_export` (#6352) Glenn Jocher 2022-01-19 12:24:40 -1000
  • e5219099cd Created using Colaboratory Glenn Jocher 2022-01-19 10:18:29 -1000
  • 0cf932bf63
    `export.py` return exported files/dirs (#6343) Glenn Jocher 2022-01-18 15:18:23 -1000
  • e2e95b2d8e
    TensorRT `assert im.device.type != 'cpu'` on export (#6340) Glenn Jocher 2022-01-18 13:52:25 -1000
  • fd55271c04
    Console corrupted -> corrupt (#6338) Glenn Jocher 2022-01-18 10:49:26 -1000
  • 3119b2f27c
    Remove `dataset_stats()` autodownload capability (#6303) Glenn Jocher 2022-01-18 10:04:49 -1000
  • db1f83be63
    Fix `train.py` parameter groups desc error (#6318) Otfot 2022-01-17 14:58:34 +0800
  • 436ffc417a
    `select_device()` cleanup (#6302) Glenn Jocher 2022-01-14 15:48:15 -1000
  • a1a9c6884c
    Fixing bug multi-gpu training (#6299) Henry 2022-01-14 21:11:06 +0100
  • e7bf38277f
    Fix `device` count check (#6290) Glenn Jocher 2022-01-13 21:23:03 -1000
  • af001349e4
    Add `is_kaggle()` function (#6285) Glenn Jocher 2022-01-13 12:39:42 -1000
  • 80473a6551
    Update `export.py` with Detect, Validate usages (#6280) Glenn Jocher 2022-01-12 15:48:40 -1000
  • f3085accd3
    Enable ONNX `--half` FP16 inference (#6268) Glenn Jocher 2022-01-11 10:13:17 -1000
  • b3eaf5008b TensorRT pip install Glenn Jocher 2022-01-10 16:49:10 -1000
  • 9b13a594e9
    Fix `cmd` string on `tfjs` export (#6243) Jinwoong Yoo 2022-01-10 15:40:47 +0900
  • 6865d19a92
    TensorRT 7 export fix (#6235) Glenn Jocher 2022-01-07 09:31:17 -0800
  • 33a67b4918
    Update P2-P7 `models/hub` variants (#6230) Glenn Jocher 2022-01-06 11:08:09 -0800
  • ad565e31d2
    Update README speed reproduction command (#6228) Glenn Jocher 2022-01-06 09:55:31 -0800
  • f80c463010
    Attempt `edgetpu-compiler` autoinstall (#6223) Glenn Jocher 2022-01-05 20:57:20 -0800
  • b4ac3df6ff
    Add `edgetpu_compiler` checks (#6218) Glenn Jocher 2022-01-05 14:55:04 -0800
  • 00d7b97869
    TFLite `--int8` 'flatbuffers==1.12' fix 2 (#6217) Glenn Jocher 2022-01-05 13:34:36 -0800
  • 8125ec5d42
    TFLite `--int8` 'flatbuffers==1.12' fix (#6216) Glenn Jocher 2022-01-05 13:01:21 -0800
  • 5402753a53
    Edge TPU compiler comment (#6196) Glenn Jocher 2022-01-04 19:36:12 -0800
  • 9e9219fe17
    Fix `nan`-robust stream FPS (#6198) Glenn Jocher 2022-01-04 19:32:42 -0800
  • b5b56a3c88
    Add CoreML inference (#6195) Glenn Jocher 2022-01-04 17:49:09 -0800
  • 7b31a531b4
    Add `tensorrt>=7.0.0` checks (#6193) Glenn Jocher 2022-01-04 13:39:13 -0800
  • a2f4a1799b
    TensorRT 7 `anchor_grid` compatibility fix (#6185) imyhxy 2022-01-05 03:09:25 +0800
  • fb839298a1
    Fix TorchScript on mobile export (#6183) Yin Rong 2022-01-04 12:25:48 +0800
  • 5bd6a97b18
    Global export format sort (#6182) Glenn Jocher 2022-01-03 20:08:15 -0800
  • 7cad6597bb
    Ignore `*_openvino_model/` dir (#6180) Glenn Jocher 2022-01-03 18:43:21 -0800
  • 63a4d862aa
    Add OpenVINO inference (#6179) Glenn Jocher 2022-01-03 15:41:26 -0800
  • b4a29b5a8d
    Update NMS `max_wh=7680` for 8k images (#6178) Glenn Jocher 2022-01-03 10:54:52 -0800
  • 5344e54da6
    [pre-commit.ci] pre-commit suggestions (#6177) pre-commit-ci[bot] 2022-01-03 10:42:50 -0800
  • 968e30065a
    Update greetings.yml (#6165) Glenn Jocher 2022-01-02 19:47:03 -0800
  • ec4b6dd2a3
    Update export format docstrings (#6151) Glenn Jocher 2022-01-02 16:09:45 -0800
  • e1dc894364
    Enable AdamW optimizer (#6152) bilzard 2022-01-03 06:10:19 +0900
  • d95978a562
    Add EdgeTPU support (#3630) Jiacong Fang 2022-01-01 01:47:52 +0800