4b2565e4cb
More forward_intermediates() / FeatureGetterNet work
...
* include relpos vit
* refactor reduction / size calcs so hybrid vits work and dynamic_img_size works
* fix -ve feature indices when pruning
* fix mvitv2 w/ class token
* refine naming
* add tests
2024-04-10 15:11:34 -07:00
d6c2cc91af
Make NormMlpClassifier head reset args consistent with ClassifierHead
2024-02-10 16:25:33 -08:00
7bc7798d0e
Type annotation correctness for create_act
2024-02-10 14:57:58 -08:00
88889de923
Fix meshgrid deprecation warnings and backward compat with explicit 'ndgrid' and 'meshgrid' fn w/o indexing arg
2024-01-27 13:48:33 -08:00
d4386219c6
Improve type handling for arange & rel pos embeds, keep calculations in float32 until application (may change to apply in float32 in future). Prevent arange type hijacking by DeepSpeed Zero
2024-01-26 16:35:51 -08:00
7f19a4cce7
fix fast catavgmax selection
2024-01-16 10:30:08 -08:00
df7ae11eb2
Add device arg for patch embed resize, fix #2024
2023-12-04 11:42:13 -08:00
9fab8d8f58
Fix break of 2 years old torchvision installs :/
2023-11-04 02:32:09 -07:00
f7762fee78
Consistency handling None / empty string inputs to norm / act create fns
2023-11-03 11:01:41 -07:00
a2e4a4c148
Add quickgelu vit clip variants, simplify get_norm_layer and allow string args in vit norm/act. Add metaclip CLIP weights
2023-11-03 11:01:41 -07:00
d5f1525334
include suggestions from review
...
Co-Authored-By: Ross Wightman <rwightman@gmail.com>
2023-10-30 13:47:54 -07:00
5f14bdd564
include typing suggestions by @rwightman
2023-10-30 13:47:54 -07:00
fe92fd93e5
fix adaptive_avgmax_pool.py
...
remove extra whitespace in `SelectAdaptivePool2d`'s `__repr__`
2023-10-29 23:03:36 -07:00
89ba0da910
Fixed Typos
...
Fixed the typos in helpers.py and CONTRIBUTING.md
2023-10-21 21:46:31 -07:00
49a459e8f1
Merge remote-tracking branch 'upstream/main' into vit_siglip_and_reg
2023-10-17 09:36:48 -07:00
a58f9162d7
Missed __init__.py update for attention pooling layer add
2023-10-17 09:28:21 -07:00
71365165a2
Add SigLIP weights
2023-10-16 23:26:08 -07:00
7ce65a83a2
Removing unused self.drop
2023-10-05 11:20:57 -07:00
9caf32b93f
Move levit style pos bias resize with other rel pos bias utils
2023-09-01 11:05:56 -07:00
170a5b6e27
add tinyvit
2023-09-01 11:05:56 -07:00
fc5d705b83
dynamic_size -> dynamic_img_size, add dynamic_img_pad for padding option
2023-08-27 15:58:35 -07:00
1f4512fca3
Support dynamic_resize in eva.py models
2023-08-27 15:58:35 -07:00
fdd8c7c2da
Initial impl of dynamic resize for existing vit models (incl vit-resnet hybrids)
2023-08-27 15:58:35 -07:00
c153cd4a3e
Add more advanced interpolation method from BEiT and support non-square window & image size adaptation for
...
* beit/beit-v2
* maxxvit/coatnet
* swin transformer
And non-square windows for swin-v2
2023-08-08 16:41:16 -07:00
76d166981d
Fix missing norm call in Mlp forward (not used by default, but can be enabled for normformer MLP scale). Fix #1851 fix #1852
2023-08-03 11:36:30 -07:00
8e4480e4b6
Patch and pos embed resample done in float32 always (cast to float and back). Fix #1811
2023-08-03 11:32:17 -07:00
40a518c194
use float in resample_abs_pos_embed_nhwc
...
since F.interpolate doesn't always support BFloat16
2023-07-28 16:01:42 -07:00
e9373b1b92
Cleanup before samvit merge. Resize abs posembed on the fly, undo some line-wraps, remove redundant unbind, fix HF hub weight load
2023-05-18 16:43:48 -07:00
a01d8f86f4
Tweak DinoV2 add, add MAE ViT weights, add initial intermediate layer getter experiment
2023-05-09 17:59:22 -07:00
5cc87e6485
Add dinov2 pretrained models ( #1797 )
...
* add dinov2 small, base, and large
* fix input size
* fix swiglu & dinov2 vit giant
* use SwiGLUPacked to replace GluMlp
* clean up & add ffn_layer placeholder for ParallelScalingBlock
2023-05-09 12:24:47 -07:00
af48246a9a
Add SwiGLUPacked to layers __init__
2023-05-08 13:52:34 -07:00
3fdb31de2e
Small SwiGLU tweak, remove default LN arg in unpacked variant, add packed alias for GluMLP
2023-05-08 12:28:00 -07:00
cb3f9c23bb
Metaformer baselines for vision (final PR with cleanup) ( #1793 )
...
* update
* Update metaformers.py
* Update metaformers.py
* Update metaformers.py
* Update metaformers.py
* Update metaformers.py
* Update metaformers.py
* Update metaformers.py
* Update metaformers.py
* Update metaformers.py
* Update metaformers.py
* Update metaformers.py
* Update metaformers.py
* Update metaformers.py
* Update metaformers.py
* Update metaformers.py
* Update metaformers.py
* Update metaformers.py
* Update metaformers.py
* Update metaformers.py
* Update metaformers.py
* Update metaformers.py
* Update metaformers.py
* Update metaformers.py
* Update metaformers.py
* Update metaformers.py
* Update metaformers.py
* Update metaformers.py
* Update metaformers.py
* Update metaformers.py
* Update metaformers.py
* Update metaformers.py
* Update metaformers.py
* Update metaformers.py
* Update metaformers.py
* Update metaformers.py
* Update metaformers.py
* Update metaformers.py
* Update metaformers.py
* Update metaformers.py
* Update metaformers.py
* Update metaformers.py
* Update metaformers.py
* Update metaformers.py
* Update metaformers.py
* merge with poolformer, initial version
* Update metaformers.py
* Update metaformers.py
* Update metaformers.py
* Update metaformers.py
* Update metaformers.py
* Update metaformers.py
* Update metaformers.py
* Update metaformers.py
* Update metaformers.py
* Update metaformers.py
* Update metaformers.py
* Update metaformers.py
* Update metaformers.py
* Update metaformers.py
* Update metaformers.py
* Update metaformers.py
* Update metaformers.py
* Update metaformers.py
* Revert "Update metaformers.py"
This reverts commit 2916f37f8d1d882eb4942209d0830e32bede4e274ed934e0003f0b0753672fef9006d7b7696a30a741fe5c3626a3aee37c35f938beb81b10bde717e539274bd45ea2329ab8ec53b8ce5b8aab6225b94102fcc30eaa366aae930426a8e481a5a913f5d438https://github.com/fffffgggg54/pytorch-image-models into metaformer_workspace
commit d577129aaa23fb348a8bb93bcd17cf1d5a4e8ff8
Author: Fredo Guan <fredo.guan@hotmail.com>
Date: Mon Mar 6 01:57:20 2023 -0800
Update metaformers.py
commit f7367304e8f3b7a9a7f16e0a032bb72546afcc2a
Author: Fredo Guan <fredo.guan@hotmail.com>
Date: Mon Mar 6 01:56:11 2023 -0800
Metaformer baselines for vision (#12 )
* formatting, cleanup, fix dropout
* fix regression, pass kwargs
* fix poolformerv1 weights, formatting
* Update metaformers.py
* Update metaformers.py
* Update metaformers.py
* Update metaformers.py
* Update metaformers.py
* Update metaformers.py
* Update metaformers.py
* Update metaformers.py
* Update metaformers.py
* Update metaformers.py
* Update metaformers.py
* Update metaformers.py
* Update metaformers.py
* some cleanup
* SDPA from ViT, fix imports
* Update metaformers.py
* fix head reset
* fast norm bias patch for metaformers
* Metaformer refactor, remove rand/ident models, fix issues, remove old poolformer
* Switch to hub weights
---------
Co-authored-by: Fredo Guan <fredo.guan@hotmail.com>
2023-05-05 11:18:26 -07:00
437d344e03
Always some torchscript issues
2023-04-26 20:42:34 -07:00
3386af8c86
Final push to get remaining models using multi-weight pretrained configs, almost all weights on HF hub
2023-04-26 15:52:13 -07:00
a08e5aed1d
More models w/ multi-weight support, moving to HF hub. Removing inplace_abn from all models including TResNet
2023-04-20 22:44:49 -07:00
aef6e562e4
Add onnx utils and export code, tweak padding and conv2d_same for better dynamic export with recent PyTorch
2023-04-11 17:03:57 -07:00
965d0a2d36
fast_attn -> fused_attn, implement global config for enable/disable fused_attn, add to more models. vit clip openai 336 weights.
2023-04-10 12:04:33 -07:00
4d135421a3
Implement patch dropout for eva / vision_transformer, refactor / improve consistency of dropout args across all vit based models
2023-04-07 20:27:23 -07:00
3863d63516
Adding EVA02 weights and model defs, move beit based eva_giant to same eva.py file. Cleanup rotary pos, add lang oriented freq bands to be compat with eva design choice. Fix #1738
2023-03-27 17:16:07 -07:00
8db20dc240
Fix #1726 , dropout not used in NormMlpClassifierHead. Make dropout more consistent across both classifier heads (nn.Dropout)
2023-03-20 09:37:05 -07:00
acfd85ad68
All swin models support spatial output, add output_fmt to v1/v2 and use ClassifierHead.
...
* update ClassifierHead to allow different input format
* add output format support to patch embed
* fix some flatten issues for a few conv head models
* add Format enum and helpers for tensor format (layout) choices
2023-03-15 23:21:51 -07:00
f77c04ff36
Torchscript fixes/hacks for rms_norm, refactor ParallelScalingBlock with manual combination of input projections, closer paper match
2023-02-16 16:57:42 -08:00
621e1b2182
Add ideas from 'Scaling ViT to 22-B Params', testing PyTorch 2.0 fused F.scaled_dot_product_attention impl in vit, vit_relpos, maxxvit / coatnet.
2023-02-16 16:57:42 -08:00
9d03c6f526
Merge remote-tracking branch 'origin/main' into levit_efficientformer_redux
2023-02-03 14:47:01 -08:00
086bd55a94
Add EfficientFormer-V2, refactor EfficientFormer and Levit for more uniformity across the 3 related arch. Add features_out support to levit conv models and efficientformer_v2. All weights on hub.
2023-02-03 14:12:29 -08:00
6f28b562c6
Factor NormMlpClassifierHead from MaxxViT and use across MaxxViT / ConvNeXt / DaViT, refactor some type hints & comments
2023-01-27 14:57:01 -08:00
ca38e1e73f
Update ClassifierHead module, add reset() method, update in_chs -> in_features for consistency
2023-01-20 14:44:05 -08:00
e520553e3e
Update batchnorm freezing to handle NormAct variants, Add GroupNorm1Act, update BatchNormAct2d tracing change from PyTorch
2023-01-12 16:55:47 -08:00
6e5553da5f
Add ConvNeXt-V2 support (model additions and weights) ( #1614 )
...
* Add ConvNeXt-V2 support (model additions and weights)
* ConvNeXt-V2 weights on HF Hub, tweaking some tests
* Update README, fixing convnextv2 tests
2023-01-05 07:53:32 -08:00
Ross Wightman
kalazus
a-r-r-o-w
Laureηt
Tush9905
lucapericlp
方曦
Sepehr Sameni
Leng Yue
Ross Wightman