Migrations
/
pytorch-image-models
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Commit Graph

  • d3f7bb4fb7 👷 update pr doc builder to bugfix branch nateraw 2022-12-05 18:21:36 -0500
  • 61b0e8296e 📝 remove augmentation from toctree nateraw 2022-12-05 17:25:06 -0500
  • d0a067705f 📝 add basic reference docs nateraw 2022-12-05 17:19:11 -0500
  • 101a7ced95 📝 wip docs nateraw 2022-12-05 17:00:11 -0500
  • e5b16e2e39 📝 wip docs nateraw 2022-12-05 16:56:12 -0500
  • ecaa5d9c9f 📝 wip docs nateraw 2022-12-05 16:34:08 -0500
  • 9c2c5bfaaa 📝 wip docs nateraw 2022-12-05 16:15:35 -0500
  • 74897ad333 📝 wip docs nateraw 2022-12-05 16:11:02 -0500
  • ffae459814 📝 add installation doc nateraw 2022-12-05 15:53:15 -0500
  • c0f8e53468 🚧 wip docs nateraw 2022-12-05 15:42:18 -0500
  • da6644b6ba Update README.md Ross Wightman 2022-12-05 10:20:08 -0800
  • 61531f091e Update multi-weight release README Ross Wightman 2022-12-05 10:16:47 -0800
  • 3785c234d7 Remove clip vit models that won't be ft and comment two that aren't uploaded yet Ross Wightman 2022-12-02 20:26:51 -0800
  • f82239b30e multi-weight branch version -> 0.8.0dev Ross Wightman 2022-12-02 17:00:56 -0800
  • 755570e2d6 Rename _pretrained.py -> pretrained.py, not feasible to change the other files to same scheme without breaking uses Ross Wightman 2022-12-02 17:00:00 -0800
  • 72cfa57761 Add ported Tensorflow MaxVit weights. Add a few more CLIP ViT fine-tunes. Tweak some model tag names. Improve model tag name sorting. Update HF hub push config layout. Ross Wightman 2022-12-02 16:54:18 -0800
  • dbe7531aa3 Update scripts to support torch.compile(). Make --results_file arg more consistent across benchmark/validate/inference. Fix #1570 Ross Wightman 2022-12-02 13:37:15 -0800
  • 05637a4bb0 More inference script changes, arg naming, multiple output fmts at once Ross Wightman 2022-11-25 23:03:54 -0800
  • eceeb9409a Significant ugprade to inference.py, support for different formats, formatting, etc. Ross Wightman 2022-11-25 17:51:38 -0800
  • 4d5c395160 MaxVit, ViT, ConvNeXt, and EfficientNet-v2 updates * Add support for TF weights and modelling specifics to MaxVit (testing ported weights) * More fine-tuned CLIP ViT configs * ConvNeXt and MaxVit updated to new pretrained cfgs use * EfficientNetV2, MaxVit and ConvNeXt high res models use squash crop/resize Ross Wightman 2022-11-21 22:14:12 -0800
  • 3db4e346e0 Switch TFDS dataset to use INTEGER_ACCURATE jpeg decode by default Ross Wightman 2022-11-21 17:42:19 -0800
  • 9da7e3a799 Add crop_mode for pretraind config / image transforms. Add support for dynamo compilation to benchmark/train/validate Ross Wightman 2022-11-21 16:30:56 -0800
  • 8fca002c06 Add ImageNet22k and 12k subset sysnet/index maps Ross Wightman 2022-11-11 09:38:51 -0800
  • b2b6285af7 Add two more FT clip weights Ross Wightman 2022-11-11 00:30:17 -0800
  • 5895056dc4 Add openai b32 ft Ross Wightman 2022-11-10 00:06:57 -0800
  • 9dea5143d5 Adding more clip ft variants Ross Wightman 2022-11-09 20:15:14 -0800
  • 444dcba4ad CLIP B16 12k weights added Ross Wightman 2022-11-09 15:27:23 -0800
  • dff4717cbf Add clip b16 384x384 finetunes Ross Wightman 2022-11-09 13:35:45 -0800
  • 883fa2eeaa Add fine-tuned B/16 224x224 in1k clip models Ross Wightman 2022-11-08 15:44:40 -0800
  • 9a3d2ac2d5 Add latest CLIP ViT fine-tune pretrained configs / model entrypt updates Ross Wightman 2022-11-07 16:02:25 -0800
  • 42bbbddee9 Add missing model config Ross Wightman 2022-11-04 12:30:45 -0700
  • def68befa7 Updating vit model defs for mult-weight support trial (vit first). Prepping for CLIP (laion2b and openai) fine-tuned weights. Ross Wightman 2022-11-01 16:23:18 -0700
  • 0dadb4a6e9 Initial multi-weight support, handled so old pretraing config handling co-exists with new tags. Ross Wightman 2022-10-26 22:18:39 -0700
  • 11704cc721
         Update README.md #1520 multi-weight Ross Wightman 2022-12-05 10:20:08 -0800
  • cfb09887fc Update multi-weight release README Ross Wightman 2022-12-05 10:16:47 -0800
  • cb725a85a2 add ViG models iamhankai 2022-12-05 23:03:58 +0800
  • 16d2db7e4b Remove clip vit models that won't be ft and comment two that aren't uploaded yet Ross Wightman 2022-12-02 20:26:51 -0800
  • 485086ae8a multi-weight branch version -> 0.8.0dev Ross Wightman 2022-12-02 17:00:56 -0800
  • c3be79a8b7 Rename _pretrained.py -> pretrained.py, not feasible to change the other files to same scheme without breaking uses Ross Wightman 2022-12-02 17:00:00 -0800
  • 23b357f1df Add ported Tensorflow MaxVit weights. Add a few more CLIP ViT fine-tunes. Tweak some model tag names. Improve model tag name sorting. Update HF hub push config layout. Ross Wightman 2022-12-02 16:54:18 -0800
  • bc8776085a Update scripts to support torch.compile(). Make --results_file arg more consistent across benchmark/validate/inference. Fix #1570 Ross Wightman 2022-12-02 13:37:15 -0800
  • 9c7e6d6907 🎨 structure docs nateraw 2022-12-02 01:02:58 -0500
  • c59d88339b Merge remote-tracking branch 'origin/main' into multi-weight Ross Wightman 2022-11-30 16:51:34 -0800
  • 3491506fec Add foreach option for faster EMA #1553 Jerome Rony 2022-11-30 14:06:58 -0500
  • e8ddc6865c
         Update torch to 1.13 and python to 3.10 in test workflow (#1562) Ross Wightman 2022-11-26 19:46:55 -0800
  • f8cbda6b55 Remove macOS from tests, not working anymore #1562 update-test-pytorch Ross Wightman 2022-11-26 14:42:20 -0800
  • 7e2e69d608 More inference script changes, arg naming, multiple output fmts at once Ross Wightman 2022-11-25 23:03:54 -0800
  • c7a07e9ee6 Significant ugprade to inference.py, support for different formats, formatting, etc. Ross Wightman 2022-11-25 17:51:38 -0800
  • f4dcf4d530 Try Python 3.10 + macOS-12 in quest for working second target Ross Wightman 2022-11-25 17:20:26 -0800
  • 98c5e78226
         Merge pull request #1564 from HongxinXiang/main Ross Wightman 2022-11-24 21:33:19 -0800
  • 8df6a5bbfd Try windows runner Ross Wightman 2022-11-24 20:01:37 -0800
  • 653bdc7105 Fix comment: https://github.com/rwightman/pytorch-image-models/pull/1564#issuecomment-1326743424 #1564 hongxin xiang 2022-11-25 09:52:52 +0800
  • f2ba9ba512 Python back to 3.9 for test workflow, see if that's the cause of macos failure.. Ross Wightman 2022-11-24 08:23:44 -0800
  • 29ae99ee16 style: trailing comma #1565 kozistr 2022-11-24 16:22:59 +0900
  • c4f9bd4fc4 update: set `indexing` parameter with `ij` kozistr 2022-11-24 16:21:46 +0900
  • bdc9fad638 Fix compatible BUG: QMNIST and ImageNet datasets do not exist in torchvision 0.10.1. hongxin xiang 2022-11-24 14:37:44 +0800
  • ef8a94ac8e Update torch and python version in test workflow Ross Wightman 2022-11-23 14:19:59 -0800
  • ce4d3485b6 refactor push_to_hub helper v0.6.12 Wauplin 2022-11-16 12:03:34 +0100
  • c037db00fc Create repo before cloning with Repository.clone_from Wauplin 2022-11-15 15:17:20 +0100
  • 3ec5918a74 Fix spacing misalignment for fast norm path in LayerNorm modules Ross Wightman 2022-10-24 21:43:49 -0700
  • 3044b1a1cb Fix xdist dependency change Ross Wightman 2022-10-26 22:32:32 -0700
  • ec6921fcb0 MaxVit, ViT, ConvNeXt, and EfficientNet-v2 updates * Add support for TF weights and modelling specifics to MaxVit (testing ported weights) * More fine-tuned CLIP ViT configs * ConvNeXt and MaxVit updated to new pretrained cfgs use * EfficientNetV2, MaxVit and ConvNeXt high res models use squash crop/resize Ross Wightman 2022-11-21 22:14:12 -0800
  • d1e0a4607d Switch TFDS dataset to use INTEGER_ACCURATE jpeg decode by default Ross Wightman 2022-11-21 17:42:19 -0800
  • 0ed0cc7eba Add crop_mode for pretraind config / image transforms. Add support for dynamo compilation to benchmark/train/validate Ross Wightman 2022-11-21 16:30:56 -0800
  • 6ec5cd6a99 Use in-place operations for EMA Jerome Rony 2022-11-17 11:53:29 -0500
  • d7165588c1 Use in-place operations for EMA #1552 Jerome Rony 2022-11-17 11:15:37 -0500
  • 25ffac6880
         Merge pull request #1547 from Wauplin/1546-fix-hf-hub-integration Ross Wightman 2022-11-16 09:19:56 -0800
  • 9b114754db refactor push_to_hub helper #1547 Wauplin 2022-11-16 12:03:34 +0100
  • ae0a0db7de Create repo before cloning with Repository.clone_from Wauplin 2022-11-15 15:17:20 +0100
  • abd4da9b9c 📌 pin hf hub version #1545 nateraw 2022-11-14 13:08:20 -0500
  • 70442fd5a8 Add ImageNet22k and 12k subset sysnet/index maps Ross Wightman 2022-11-11 09:38:51 -0800
  • f6e0a848d0 Add two more FT clip weights Ross Wightman 2022-11-11 00:30:17 -0800
  • 884a0f1a12 Add openai b32 ft Ross Wightman 2022-11-10 00:06:57 -0800
  • 2c80da3b9a Adding more clip ft variants Ross Wightman 2022-11-09 20:15:14 -0800
  • da2de0de95 CLIP B16 12k weights added Ross Wightman 2022-11-09 15:27:23 -0800
  • b2897f5ea6 Add clip b16 384x384 finetunes Ross Wightman 2022-11-09 13:35:45 -0800
  • 092287436e Add fine-tuned B/16 224x224 in1k clip models Ross Wightman 2022-11-08 15:44:40 -0800
  • 8ff45e41f7 Fix xdist dependency change Ross Wightman 2022-10-26 22:32:32 -0700
  • d4ed3e6181
         Update pr docs actions #1538 Mishig 2022-11-08 16:20:08 +0100
  • 1ba6902461
         Merge 2be436cb2c into 4714a4910e #1210 Daniel Nouri 2022-11-08 12:59:40 +0900
  • d3415e3134 Add latest CLIP ViT fine-tune pretrained configs / model entrypt updates Ross Wightman 2022-11-07 16:02:25 -0800
  • 2eb825c014 Add missing model config Ross Wightman 2022-11-04 12:30:45 -0700
  • 4714a4910e
         Merge pull request #1525 from TianyiFranklinWang/main Ross Wightman 2022-11-03 20:55:43 -0700
  • f856f8a2e0
         Merge pull request #1527 from klae01/patch-1 Ross Wightman 2022-11-03 20:54:59 -0700
  • 7b8a839b24 fix train args 'in_chanEs' typo #1532 Raul Almeida 2022-11-03 10:08:53 -0300
  • a25bf974a9
         Update README.md #804 bits_and_tpu Ross Wightman 2022-11-01 21:49:10 -0700
  • 0761ce7a1b Updating vit model defs for mult-weight support trial (vit first). Prepping for CLIP (laion2b and openai) fine-tuned weights. Ross Wightman 2022-11-01 16:23:18 -0700
  • ddd6361904
         Update train.py #1527 klae01 2022-11-01 16:55:05 +0900
  • 9152b10478
         ✏️ fix typo #1525 NPU-Franklin 2022-10-30 08:49:40 +0800
  • 4201c718e9
         Merge pull request #1522 from hova88/main Ross Wightman 2022-10-27 21:02:35 -0700
  • 29baf32327 fix typo : miss back quote #1522 hova88 2022-10-28 09:30:51 +0800
  • 041709b470 Fix xdist dependency change Ross Wightman 2022-10-26 22:32:32 -0700
  • ebb99a1f8d Initial multi-weight support, handled so old pretraing config handling co-exists with new tags. Ross Wightman 2022-10-26 22:18:39 -0700
  • 803254bb40 Fix spacing misalignment for fast norm path in LayerNorm modules Ross Wightman 2022-10-24 21:43:49 -0700
  • 475ecdfa3d cast env var args for dataset readers to int Ross Wightman 2022-10-17 14:40:11 -0700
  • 44cea489fb
         Merge pull request #1502 from infomon/main Ross Wightman 2022-10-17 14:06:23 -0700
  • aceb79e002 Fix typo #1502 Simon Schrodi 2022-10-17 22:06:17 +0200
  • 8a1ba67a50
         Merge pull request #1500 from nguyenhoan1988/main Ross Wightman 2022-10-17 09:05:04 -0700
  • 39190f5f44
         Remove inplace operators when calculating the loss #1500 Hoan Nguyen 2022-10-17 10:41:21 +0200
  • 91fc4c7ed1
         Update tests.yml Ross Wightman 2022-10-16 20:51:20 -0700