Commit Graph

1613 Commits (bits_and_tpu)
 

Author SHA1 Message Date
Ross Wightman 4b30bae67b Add updated vit_relpos weights, and impl w/ support for official swin-v2 differences for relpos. Add bias control support for MLP layers
2 years ago
Ross Wightman d4c0588012 Remove persistent buffers from Swin-V2. Change SwinV2Cr cos attn + tau/logit_scale to match official, add ckpt convert, init_value zeros resid LN weight by default
2 years ago
Ross Wightman 27c42f0830 Fix torchscript use for offician Swin-V2, add support for non-square window/shift to WindowAttn/Block
2 years ago
Ross Wightman 2f2b22d8c7 Disable nvfuser fma / opt level overrides per #1244
2 years ago
Ross Wightman c0211b0bf7 Swin-V2 test fixes, typo
2 years ago
Ross Wightman 9a86b900fa Official SwinV2 models
2 years ago
Ross Wightman d07d015173
Merge pull request #1249 from okojoalg/sequencer
2 years ago
Ross Wightman d30685c283
Merge pull request #1251 from hankyul2/fix-multistep-scheduler
2 years ago
han a16171335b fix: change milestones to decay-milestones
2 years ago
Ross Wightman 39b725e1c9 Fix tests for rank-4 output where feature channels dim is -1 (3) and not 1
2 years ago
Ross Wightman 78a32655fa Fix poolformer group_matcher to merge proj downsample with previous block, support coarse
2 years ago
Ross Wightman d79f3d9d1e Fix torchscript use for sequencer, add group_matcher, forward_head support, minor formatting
2 years ago
Ross Wightman 37b6920df3 Fix group_matcher regex for regnet.py
2 years ago
okojoalg 93a79a3dd9 Fix num_features in Sequencer
2 years ago
okojoalg 2fec08e923 Add Sequencer to non std filters
2 years ago
han 57a988df30 fix: multistep lr decay epoch bugs
2 years ago
okojoalg 578d52e752 Add Sequencer
2 years ago
Ross Wightman 6d4665bb52
Merge pull request #1245 from rwightman/vit_relpos_refactor
2 years ago
Ross Wightman f5ca4141f7 Adjust arg order for recent vit model args, add a few comments
2 years ago
Ross Wightman 41dc49a337 Vision Transformer refactoring and Rel Pos impl
2 years ago
Ross Wightman 9c321be330
Merge pull request #1239 from dedeswim/parser-fix
2 years ago
Edoardo Debenedetti c76d772670 Add support for different TFDS `BuilderConfig`s
2 years ago
Ross Wightman b7cb8d0337 Add Swin-V2 Small-NS weights (83.5 @ 224). Add layer scale like 'init_values' via post-norm LN weight scaling
2 years ago
Ross Wightman 001688dabf
Merge pull request #1233 from jjsjann123/nhwc_cond_conv2d
2 years ago
jjsjann123 f88c606fcf fixing channels_last on cond_conv2d; update nvfuser debug env variable
2 years ago
Ross Wightman 7d235c5a5f
Merge pull request #1230 from donglixp/patch-1
2 years ago
Li Dong 09e9f3defb
migrate azure blob for beit checkpoints
2 years ago
Ross Wightman 52ac881402 Missed first_conv in latest seresnext 'D' default_cfgs
2 years ago
Ross Wightman 7629d8264d Add two new SE-ResNeXt101-D 32x8d weights, one anti-aliased and one not. Reshuffle default_cfgs vs model entrypoints for resnet.py so they are better aligned.
2 years ago
Ross Wightman fbf597049c Update README and change timmdocs link in documentation
2 years ago
Ross Wightman 01a0e25a67
Merge pull request #1208 from seefun/master
2 years ago
SeeFun 8f0bc0591e fix convnext args
2 years ago
SeeFun b0d2fcf647
Merge branch 'rwightman:master' into master
2 years ago
Ross Wightman 754e11402a Merge remote-tracking branch 'origin/master' into bits_and_tpu
2 years ago
Ross Wightman eac2df3d2c Update PyTorch 1.10 benchmark numbers for latest code
2 years ago
Ross Wightman c5a8e929fb Add initial swinv2 tiny / small weights
2 years ago
Ross Wightman 83d7a11eec Update .gitignore, remove out of date notebooks
2 years ago
Ross Wightman 02b806e00a Add PyTorch 1.11 train benchmark numbers
2 years ago
Ross Wightman c9f208f7f8 Update eval results and add latest PyTorch 1.11 inference benchmarks
2 years ago
Ross Wightman 1ba0ec4c18 Merge remote-tracking branch 'origin/master' into bits_and_tpu
2 years ago
Ross Wightman f670d98cb8 Make a few more layers symbolically traceable (remove from FX leaf modules)
2 years ago
Ross Wightman a9ecb880e5
Merge pull request #1190 from seefun/ConvNeXt-pretrain
2 years ago
SeeFun 5f4de2334b
Merge pull request #1 from seefun/ConvNeXt-pretrain
2 years ago
SeeFun ec4e9aa5a0
Add ConvNeXt tiny and small pretrain in22k
2 years ago
Ross Wightman 73ffade1f8
Update README.md
2 years ago
Ross Wightman 575924ed60 Update test crop for new RegNet-V weights to match Y
2 years ago
Ross Wightman 8f6d638887 Update README.md
2 years ago
Ross Wightman 1618527098 Add layer scale and parallel blocks to vision_transformer
2 years ago
Ross Wightman c42be74621 Add attrib / comments about Swin-S3 (AutoFormerV2) weights
2 years ago
Ross Wightman 474ac906a2 Add 'head norm first' convnext_tiny_hnf weights
2 years ago