Commit Graph

Select branches

Hide Pull Requests

0.5.x

0.6.x

attn_update

bits_and_tpu

cleanup_xla_model_fixes

clip_laion2b

convnext_and_copyright

convnextv2

dataset_info

edgenext_csp_and_more

efficientnet_attn

eva

fix_tests

fix_window_trace

fixes-syncbn_pretrain_cfg_resolve

fixes_bce_regnet

focalnet_and_swin_refactor

freeze_norm_revisit

gh-pages

hf_hub_revisit

levit_efficientformer_redux

main

maxxvit_hub

more_attn

more_datasets

more_vit

multi-weight

multi-weight_effnet_convnext

norm_mlp_classifier

norm_norm_norm

parse_kwargs

parse_kwargs2

perceiver

refactor-imports

scaling_vit

script_cleanup

swin_v2

update-test-pytorch

vit_and_bit_test_fixes

vit_relpos_refactor

xsepconv

#1

#1007

#101

#1012

#1014

#1034

#1037

#105

#1055

#1056

#1056

#1058

#1061

#1063

#1063

#1069

#1083

#1091

#1094

#1105

#1112

#1113

#115

#1150

#117

#1187

#1190

#1194

#12

#1208

#1210

#1210

#1212

#1213

#1213

#122

#1222

#1222

#1229

#123

#1230

#1233

#1239

#1245

#1249

#125

#1251

#1253

#1253

#1256

#1259

#1265

#1266

#1270

#1271

#1271

#1273

#1274

#1276

#1277

#1294

#13

#1317

#132

#1322

#1327

#1330

#1330

#1336

#1340

#1345

#1349

#1351

#1354

#136

#1363

#1365

#137

#1381

#1383

#1385

#140

#141

#1414

#1415

#142

#1420

#143

#1439

#1439

#145

#146

#1467

#1471

#1473

#1476

#1479

#148

#1483

#1493

#1494

#1498

#150

#1500

#1502

#1520

#1522

#1525

#1527

#1532

#1538

#154

#1545

#1547

#155

#1552

#1553

#1553

#156

#1562

#1564

#1565

#1575

#1578

#1578

#1581

#1582

#1583

#1586

#1590

#1590

#1592

#1593

#16

#1606

#1612

#1614

#1617

#1617

#1621

#1622

#1624

#1627

#1628

#1628

#163

#1630

#1633

#1641

#1643

#1643

#1645

#1647

#1647

#1649

#1654

#1654

#1655

#1659

#1659

#1662

#1673

#1674

#1677

#1679

#1679

#1680

#1683

#1685

#1687

#1688

#1688

#175

#179

#183

#19

#191

#195

#199

#2

#208

#212

#212

#213

#213

#214

#216

#218

#227

#228

#23

#233

#237

#24

#240

#240

#244

#250

#255

#256

#263

#268

#282

#284

#286

#29

#290

#290

#297

#30

#300

#302

#31

#315

#32

#322

#323

#33

#334

#335

#35

#352

#367

#382

#389

#401

#413

#418

#419

#421

#425

#427

#429

#434

#437

#440

#447

#450

#459

#475

#489

#494

#497

#501

#502

#510

#52

#525

#528

#53

#533

#537

#542

#548

#550

#554

#555

#556

#561

#571

#580

#581

#595

#6

#603

#603

#609

#612

#613

#62

#625

#626

#636

#637

#645

#647

#65

#651

#652

#653

#659

#659

#66

#660

#668

#669

#679

#681

#685

#693

#7

#702

#714

#729

#731

#738

#74

#746

#747

#750

#768

#768

#771

#772

#800

#801

#804

#804

#805

#808

#812

#813

#816

#817

#82

#821

#828

#83

#841

#842

#853

#853

#858

#87

#874

#875

#876

#88

#880

#882

#891

#898

#910

#913

#914

#915

#933

#94

#95

#964

#967

#968

#986

#989

#99

#993

v0.1-attn-weights

v0.1-cadene

v0.1-coat-weights

v0.1-dnf-weights

v0.1-effv2-weights

v0.1-ger-weights

v0.1-hrnet

v0.1-mvit-weights

v0.1-pit-weights

v0.1-regnet

v0.1-repvgg-weights

v0.1-res2net

v0.1-resnest

v0.1-rexnet

v0.1-rs-weights

v0.1-rsb-weights

v0.1-selecsls

v0.1-tpu-weights

v0.1-tresnet

v0.1-vitjx

v0.1-vt3p-weights

v0.1-weights

v0.1-weights-maxx

v0.1-weights-morevit

v0.1-weights-swinv2

v0.1.1

v0.2.1

v0.3.3

v0.3.4

v0.4.12

v0.4.5

v0.4.9

v0.5.4

v0.6.11

v0.6.12

v0.6.5

v0.6.7

v0.8.10dev0

v0.8.13dev0

v0.8.2dev0

v0.8.6dev0

Mono Color

• acc3ed2b8c Add EfficientNet-B3 weights, trained from scratch with RA. Ross Wightman 2019-12-28 17:24:15 -0800
• 2404361f62 correct asset paths #65 Dushyant Mehta 2019-12-28 23:32:20 +0100
• 31939311f6 Added SelecSLS models Dushyant Mehta 2019-12-28 23:06:00 +0100
• 1f4498f217 Add ResNet deep tiered stem and model weights for seresnext26t_32x4d and seresnext26d_32x4d rwightman 2019-12-28 11:42:36 -0800
• 32012a44fd Added SelecSLS model Dushyant Mehta 2019-12-28 20:41:55 +0100
• 73b78459dc Add update RandAugment MixNet-XL weights Ross Wightman 2019-12-24 10:08:24 -0800
• 3afc2a4dc0 Some cleanup/improvements to AugMix impl: * make 'increasing' levels for Contrast, Color, Brightness, Saturation ops * remove recursion from faster blending mix * add config striing parsing for AugMix Ross Wightman 2019-12-20 22:32:05 -0800
• 232ab7fb12 Working on an implementation of AugMix with JensenShannonDivergence loss that's compatible with my AutoAugment and RandAugment impl Ross Wightman 2019-12-19 18:16:18 -0800
• ff8688ca3d
  Merge pull request #62 from rwightman/reduce-bn Ross Wightman 2019-12-19 23:22:16 -0800
• 5719b493ad Missed update dist-bn logic for EMA model #62 Ross Wightman 2019-12-19 23:03:04 -0800
• a435ea1327 Change reduce_bn to distribute_bn, add ability to choose between broadcast and reduce (mean). Add crop_pct arg to allow selecting validation crop while training. Ross Wightman 2019-12-19 22:56:54 -0800
• 3bff2b21dc Add support for keeping running bn stats the same across distributed training nodes before eval/save Ross Wightman 2019-12-05 22:35:40 -0800
• 0161de0127 Switch RandoErasing back to on GPU normal sampling Ross Wightman 2019-12-05 22:35:08 -0800
• 5d7af97981
  Update README.md Ross Wightman 2019-12-05 11:47:57 -0800
• 3129bdb2c1
  Update README.md Ross Wightman 2019-12-04 11:13:25 -0800
• ff421e5e09 New PyTorch trained EfficientNet-B2 weights with my RandAugment impl Ross Wightman 2019-12-04 11:09:47 -0800
• 00b93407c7 Update results-all.csv with latest models/weights Ross Wightman 2019-11-29 23:25:53 -0800
• 5259dbcbb4 Update README.md with latest changes Ross Wightman 2019-11-29 17:45:35 -0800
• 3bef524f9c Finish with HRNet, weights and models updated. Improve consistency in model classifier/global pool treatment. Ross Wightman 2019-11-29 17:56:36 -0800
• 3ceeedc441
  Merge pull request #53 from rwightman/condconvs_and_features v0.1-hrnet Ross Wightman 2019-11-28 18:05:32 -0800
• 7b3c235ccc Update sotabench with new models #53 Ross Wightman 2019-11-28 17:56:59 -0800
• 6ca0828166 Update EfficientNet comments, MobileNetV3 non-TF create fns, fix factory arg checks, bump PyTorch version req to 1.2 Ross Wightman 2019-11-28 17:11:53 -0800
• eccbadca74 Update EfficientNet comments Ross Wightman 2019-11-28 17:11:53 -0800
• 902d32fb16 Renamed gen_efficientnet.py -> efficientnet.py Ross Wightman 2019-11-28 16:55:36 -0800
• 5a0a8de7e3 ResNet updates: * remove redundant GluonResNet model/blocks and use the code in ResNet for Gluon weights * change SEModules back to using AdaptiveAvgPool instead of mean, PyTorch issue long fixed Ross Wightman 2019-11-28 16:46:21 -0800
• 3b8f63084d Clean checkpoint renames pth w/ SHA hash in a torch hub compatible way Ross Wightman 2019-11-28 16:42:37 -0800
• a39cc43374 Bring EfficientNet and MobileNetV3 up to date with my gen-efficientnet repo * Split MobileNetV3 and EfficientNet model files and put builder and blocks in own files (getting too large) * Finalize CondConv EfficientNet variant * Add the AdvProp weights files and B8 EfficientNet model * Refine the feature extraction module for EfficientNet and MobileNetV3 Ross Wightman 2019-11-28 16:40:35 -0800
• ad93347548 Initial HRNet classification model commit Ross Wightman 2019-11-25 22:34:04 -0800
• 2393708650 Missed stashing of out_indices in model Ross Wightman 2019-11-05 09:55:55 -0800
• 35e8f0c5e7 Fixup a few comments, add PyTorch version aware Flatten and finish as_sequential for GenEfficientNet Ross Wightman 2019-10-23 11:37:03 -0700
• 7ac6db4543 Missed activations.py Ross Wightman 2019-10-22 23:45:30 -0700
• 506df0e3d0 Add CondConv support for EfficientNet into WIP for GenEfficientNet Feature extraction setup Ross Wightman 2019-10-22 23:42:04 -0700
• db04677c94
  Merge pull request #52 from rwightman/randaugment Ross Wightman 2019-11-22 14:05:29 -0800
• 576d360f20 Bring in JIT version of optimized swish activation from gen_efficientnet as default (while working on feature extraction functionality here). #52 Ross Wightman 2019-11-22 13:57:45 -0800
• 1f39d15f15 Allow float decay epochs arg for training, works out with step lr math Ross Wightman 2019-11-22 13:28:24 -0800
• 7b83e67f77 Pass drop connect arg through to EfficientNet models Ross Wightman 2019-11-22 13:27:43 -0800
• 31453b039e Update Auto/RandAugment comments, README, more. * Add a weighted choice option for RandAugment * Adjust magnitude noise/std naming, config Ross Wightman 2019-11-22 13:00:24 -0800
• 4243f076f1 Adding RandAugment to AutoAugment impl, some tweaks to AA included Ross Wightman 2019-11-21 21:14:33 -0800
• 4748c6dff2 Fix non-prefetch variant of Mixup. Fixes #50 Ross Wightman 2019-11-02 17:42:55 -0700
• 0d58c50fb1 Add TF RandAug weights for B5/B7 EfficientNet models. Ross Wightman 2019-10-30 16:47:39 -0700
• 62105ed4a0 Better differentiate sotabench WSL, SSL, and SWSL models via model_desc Ross Wightman 2019-10-20 13:17:25 -0700
• c099374771 Map pretrained checkpoint to cpu to avoid issue with some pretrained checkpoints still having CUDA tensors. Fixes #42 Ross Wightman 2019-10-19 17:27:46 -0700
• b93fcf0708 Add Facebook Research Semi-Supervised and Semi-Weakly Supervised ResNet model weights. Ross Wightman 2019-10-19 17:05:37 -0700
• a9eb484835 Add memory efficient Swish impl Ross Wightman 2019-10-19 14:48:30 -0700
• 187ecbafbe Add support for loading args from yaml file (and saving them with each experiment) Ross Wightman 2019-09-09 17:56:45 -0700
• d3ba34ee7e Fix Mobilenet V3 model name for sotabench. Minor res2net cleanup. rwightman 2019-09-05 15:47:56 -0700
• b5a8bb52fd Silly typos. rwightman 2019-09-04 21:36:58 -0700
• 7dc5d7a997 sotabench fail rwightman 2019-09-04 19:15:36 -0700
• 96364fc336
  Merge pull request #35 from rwightman/res2net_dla Ross Wightman 2019-09-04 17:41:50 -0700
• 2680ad14bb Add Res2Net and DLA to README #35 Ross Wightman 2019-09-04 17:38:59 -0700
• adbf770f16 Add Res2Net and DLA models w/ pretrained weights. Update sotabench. rwightman 2019-09-04 17:06:42 -0700
• 7fd0857bfe
  Merge pull request #33 from rwightman/autoaugment v0.1-res2net Ross Wightman 2019-09-01 22:10:03 -0700
• 4002c0d4ce Fix AutoAugment abs translate calc #33 Ross Wightman 2019-09-01 22:07:45 -0700
• c06274e5a2 Add note on random selection of magnitude value Ross Wightman 2019-09-01 20:32:26 -0700
• b750b76f67 More AutoAugment work. Ready to roll... Ross Wightman 2019-09-01 16:55:42 -0700
• 25d2088d9e Working on auto-augment Ross Wightman 2019-08-31 23:09:48 -0700
• aff194f42c
  Merge pull request #32 from rwightman/opt Ross Wightman 2019-08-29 15:26:15 -0700
• 64966f61f7 Add Nvidia's NovogGrad impl from Jasper (cleaner/faster than current) and Apex Fused optimizers #32 Ross Wightman 2019-08-29 15:21:38 -0700
• 3d9c8a6489 Add support for new AMP checkpointing support w/ amp.state_dict Ross Wightman 2019-08-29 15:19:18 -0700
• ba3c97c3ad Some Lookahead cleanup and fixes Ross Wightman 2019-08-29 15:14:35 -0700
• 5c6da1c55a Experiment with sotabench model_description field. Ross Wightman 2019-08-28 21:34:39 -0700
• e9d2ec4d8e
  Merge pull request #31 from rwightman/opt Ross Wightman 2019-08-28 00:20:39 -0700
• fac58f609a Add RAdam, NovoGrad, Lookahead, and AdamW optimizers, a few ResNet tweaks and scheduler factory tweak. * Add some of the trendy new optimizers. Decent results but not clearly better than the standards. * Can create a None scheduler for constant LR * ResNet defaults to zero_init of last BN in residual * add resnet50d config #31 Ross Wightman 2019-08-28 00:14:10 -0700
• 81875d52a6 Update sotabench model list, add Mean-Max pooling DPN variants, disable download progress Ross Wightman 2019-08-27 00:07:32 -0700
• f37e633e9b Merge remote-tracking branch 'origin/re-exp' into opt Ross Wightman 2019-08-26 14:29:23 -0700
• b06dce8d71 Bump version for next push to pypi Ross Wightman 2019-08-25 22:32:12 -0700
• 32a0ca03c8 Update csv results files Ross Wightman 2019-08-25 22:27:14 -0700
• dc0630fd94
  Merge pull request #30 from rwightman/mixnet_xl Ross Wightman 2019-08-24 19:59:11 -0700
• 73fbd97ed4 Add weights for my MixNet-XL creation, include README updates for EdgeTPU models #30 Ross Wightman 2019-08-24 19:52:01 -0700
• 51a2375b0c Experimenting with a custom MixNet-XL and MixNet-XXL definition Ross Wightman 2019-08-12 15:01:00 -0700
• 190895fc84 Add weights for my MixNet-XL creation, include README updates for EdgeTPU models #29 Ross Wightman 2019-08-24 19:52:01 -0700
• 54cdf892b0 Add EdgeTPU models to sotabench, remove some unicode chars from model names Ross Wightman 2019-08-24 14:42:51 -0700
• 2a919886e2 Finally got around to adding EdgeTPU EfficientNet variant Ross Wightman 2019-08-24 13:39:16 -0700
• 25365c99e1 Add initial sotabench attempt. Split create_transform out of create_loader. Update requirements.txt Ross Wightman 2019-08-12 23:06:19 -0700
• ce66269dba A few more additions to Gluon Xception models to match interface of others. Ross Wightman 2019-08-11 23:06:23 -0700
• 12c954c9b0 Update helpers.py Minqin Chen 2019-08-11 04:21:39 +0800
• 9816ca3ab4 Add EdgeTPU models to sotabench, remove some unicode chars from model names Ross Wightman 2019-08-24 14:42:51 -0700
• 9ec6824bab Finally got around to adding EdgeTPU EfficientNet variant Ross Wightman 2019-08-24 13:39:16 -0700
• daeaa113e2 Add initial sotabench attempt. Split create_transform out of create_loader. Update requirements.txt Ross Wightman 2019-08-12 23:06:19 -0700
• 66634d2200 Add support to split random erasing blocks into randomly selected number with --recount arg. Fix random selection of aspect ratios. Ross Wightman 2019-08-12 15:53:40 -0700
• 6946281fde Experimenting with random erasing changes Ross Wightman 2019-07-28 10:27:26 -0700
• 2cd2162202 Experimenting with a custom MixNet-XL and MixNet-XXL definition Ross Wightman 2019-08-12 15:01:00 -0700
• aeaaad7304
  Merge pull request #24 from rwightman/gluon_xception Ross Wightman 2019-08-11 23:08:21 -0700
• 3b4868f6dc A few more additions to Gluon Xception models to match interface of others. #24 Ross Wightman 2019-08-11 23:06:23 -0700
• f937e2a600
  Merge pull request #23 from aovoc/patch-1 Ross Wightman 2019-08-10 15:47:40 -0700
• 4d505e0785 Add working Gluon Xception-65 model. Some cleanup still needed. Ross Wightman 2019-08-10 13:52:01 -0700
• 4e7a854dd0
  Update helpers.py #23 Minqin Chen 2019-08-11 04:21:39 +0800
• 0c874195db Update results csv files, bump version for timm pip release Ross Wightman 2019-08-05 11:32:49 -0700
• 9b35818811
  Update README.md Ross Wightman 2019-08-03 09:32:08 -0700
• 4fe2da558c Add MixNet Small and Large PyTorch native weights (no same padding) Ross Wightman 2019-08-02 23:16:07 -0700
• 15e66863b1
  Update README.md Ross Wightman 2019-07-31 14:41:31 -0700
• e879cf52fa Update validation scores for new TF EfficientNet weights. Ross Wightman 2019-07-31 14:36:38 -0700
• b35050a144
  Update README.md Ross Wightman 2019-07-30 18:33:40 -0700
• 77e2e0c4e3 Add new auto-augmentation Tensorflow EfficientNet weights, incl B6 and B7 models. Validation scores still pending but looking good. Ross Wightman 2019-07-30 18:31:02 -0700
• 857f33015a Add native PyTorch weights for MixNet-Medium with no SAME padding necessary. Remove unused block of code. Ross Wightman 2019-07-29 11:58:13 -0700
• e7c8a37334 Make min-lr and cooldown-epochs cmdline args, change dash in color_jitter arg for consistency Ross Wightman 2019-07-26 09:35:31 -0700
• d4debe6597 Update version, results csv files, and move remaining dropbox weights to github Ross Wightman 2019-07-25 16:54:44 -0700
• dfa9298b4e Add MixNet (https://arxiv.org/abs/1907.09595) with pretrained weights converted from Tensorflow impl * refactor 'same' convolution and add helper to use MixedConv2d when needed * improve performance of 'same' padding for cases that can be handled statically * add support for extra exp, pw, and dw kernel specs with grouping support to decoder/string defs for MixNet * shuffle some args for a bit more consistency, a little less clutter overall in gen_efficientnet.py Ross Wightman 2019-07-25 11:38:05 -0700
• 7a92caa560 Add basic image folder style dataset to read directly out of tar files, example in validate.py Ross Wightman 2019-07-25 10:51:03 -0700
• d6ac5bbc48 EfficientNet and related cleanup * remove folded_bn support and corresponding untrainable tflite ported weights * combine bn args into dict * add inplace support to activations and use where possible for reduced mem on large models Ross Wightman 2019-07-22 09:25:52 -0700