You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.

Ross Wightman 9709dbaaa9 Adding support for fine-tune CLIP LAION-2B image tower weights for B/32, L/14, H/14 and g/14. Still WIP		2 years ago
..
__init__.py	MaxxVit! A very configurable MaxVit and CoAtNet impl with lots of goodies..	2 years ago
activations.py	…
activations_jit.py	…
activations_me.py	…
adaptive_avgmax_pool.py	Refactoring, cleanup, improved test coverage.	3 years ago
attention_pool2d.py	Significant model refactor and additions:	3 years ago
blur_pool.py	Make a few more layers symbolically traceable (remove from FX leaf modules)	3 years ago
bottleneck_attn.py	wip - pre-rebase	3 years ago
cbam.py	Significant norm update	3 years ago
classifier.py	Significant model refactor and additions:	3 years ago
cond_conv2d.py	fixing channels_last on cond_conv2d; update nvfuser debug env variable	3 years ago
config.py	…
conv2d_same.py	…
conv_bn_act.py	Add more experimental darknet and 'cs2' darknet variants (different cross stage setup, closer to newer YOLO backbones) for train trials.	2 years ago
create_act.py	Add 'fast' layer norm that doesn't cast to float32, support APEX LN impl for slight speed gain, update norm and act factories, tweak SE for ability to disable bias (needed by GCVit)	2 years ago
create_attn.py	Add MobileVitV2 support. Fix #1332. Move GroupNorm1 to common layers (used in poolformer + mobilevitv2). Keep ol custom ConvNeXt LayerNorm2d impl as LayerNormExp2d for reference.	2 years ago
create_conv2d.py	Significant norm update	3 years ago
create_norm.py	Add 'fast' layer norm that doesn't cast to float32, support APEX LN impl for slight speed gain, update norm and act factories, tweak SE for ability to disable bias (needed by GCVit)	2 years ago
create_norm_act.py	More model and test fixes	2 years ago
drop.py	Improve repr for DropPath module	2 years ago
eca.py	Add non-local and BAT attention. Merge attn and self-attn factories into one. Add attention references to README. Add mlp 'mode' to ECA.	4 years ago
evo_norm.py	Remove SiLU layer in default args that breaks import on old old PyTorch	2 years ago
fast_norm.py	Add norm/norm_act header comments	2 years ago
filter_response_norm.py	Updated EvoNorm implementations with some experimentation. Add FilterResponseNorm. Updated RegnetZ and ResNetV2 model defs for trials.	3 years ago
gather_excite.py	Add Gather-Excite and Global Context attn modules. Refactor existing SE-like attn for consistency and refactor byob/byoanet for less redundancy.	4 years ago
global_context.py	wip - pre-rebase	3 years ago
halo_attn.py	wip - pre-rebase	3 years ago
helpers.py	MaxxVit! A very configurable MaxVit and CoAtNet impl with lots of goodies..	2 years ago
inplace_abn.py	Significant norm update	3 years ago
lambda_layer.py	wip - pre-rebase	3 years ago
linear.py	…
median_pool.py	…
mixed_conv2d.py	…
ml_decoder.py	support CNNs	3 years ago
mlp.py	Add updated vit_relpos weights, and impl w/ support for official swin-v2 differences for relpos. Add bias control support for MLP layers	3 years ago
non_local_attn.py	Significant norm update	3 years ago
norm.py	Add norm/norm_act header comments	2 years ago
norm_act.py	Add norm/norm_act header comments	2 years ago
padding.py	…
patch_embed.py	Adding support for fine-tune CLIP LAION-2B image tower weights for B/32, L/14, H/14 and g/14. Still WIP	2 years ago
pool2d_same.py	Support npz custom load for vision transformer hybrid models. Add posembed rescale for npz load.	3 years ago
pos_embed.py	Significant model refactor and additions:	3 years ago
selective_kernel.py	Significant norm update	3 years ago
separable_conv.py	Significant norm update	3 years ago
space_to_depth.py	…
split_attn.py	Significant norm update	3 years ago
split_batchnorm.py	…
squeeze_excite.py	Add 'fast' layer norm that doesn't cast to float32, support APEX LN impl for slight speed gain, update norm and act factories, tweak SE for ability to disable bias (needed by GCVit)	2 years ago
std_conv.py	Use reshape instead of view in std_conv, causing issues in recent PyTorch in channels_last	3 years ago
test_time_pool.py	Add --use-train-size flag to force use of train input_size (over test input size) for validation. Default test-time pooling to use train input size (fixes issues).	2 years ago
trace_utils.py	Better fix for #954 that doesn't break torchscript, pull torch._assert into timm namespace when it exists	3 years ago
weight_init.py	Clean up no_grad for trunc normal weight inits	2 years ago