Ross Wightman 30ffa152de Fix load of larger ResNet CLIP models, experimenting with making AttentionPool *the* head, seems to fine-tune better, one less layer.

2024-06-10 12:07:14 -07:00

..

__init__.py

Add anti-aliasing support to mobilenetv3 and efficientnet family models. Update MobileNetV4 model defs, resolutions. Fix #599

2024-05-27 22:06:22 -07:00

activations_me.py

Remove JIT activations, take jit out of ME activations. Remove other instances of torch.jit.script. Breaks torch.compile and is much less performant. Remove SpaceToDepthModule

2024-05-06 16:32:49 -07:00

activations.py

…

adaptive_avgmax_pool.py

…

attention2d.py

More MobileNet-v4 fixes

2024-05-24 15:09:29 -07:00

attention_pool2d.py

Fix load of larger ResNet CLIP models, experimenting with making AttentionPool *the* head, seems to fine-tune better, one less layer.

2024-06-10 12:07:14 -07:00

attention_pool.py

Mapping OpenAI CLIP Modified ResNet weights -> ByobNet. Improve AttentionPool2d layers. Fix #1731

2024-06-09 16:54:48 -07:00

blur_pool.py

Add anti-aliasing support to mobilenetv3 and efficientnet family models. Update MobileNetV4 model defs, resolutions. Fix #599

2024-05-27 22:06:22 -07:00

bottleneck_attn.py

…

cbam.py

…

classifier.py

Mapping OpenAI CLIP Modified ResNet weights -> ByobNet. Improve AttentionPool2d layers. Fix #1731

2024-06-09 16:54:48 -07:00

cond_conv2d.py

…

config.py

…

conv2d_same.py

…

conv_bn_act.py

Remove separate ConvNormActAa class, merge with ConvNormAct

2024-06-10 12:05:35 -07:00

create_act.py

Remove JIT activations, take jit out of ME activations. Remove other instances of torch.jit.script. Breaks torch.compile and is much less performant. Remove SpaceToDepthModule

2024-05-06 16:32:49 -07:00

create_attn.py

…

create_conv2d.py

…

create_norm_act.py

…

create_norm.py

Move norm & pool into Hiera ClassifierHead. Misc fixes, update features_intermediate() naming

2024-05-11 23:37:35 -07:00

drop.py

…

eca.py

…

evo_norm.py

…

fast_norm.py

…

filter_response_norm.py

…

format.py

…

gather_excite.py

…

global_context.py

…

grid.py

…

grn.py

…

halo_attn.py

…

helpers.py

…

inplace_abn.py

…

interpolate.py

…

lambda_layer.py

…

linear.py

…

median_pool.py

…

mixed_conv2d.py

…

ml_decoder.py

Remove JIT activations, take jit out of ME activations. Remove other instances of torch.jit.script. Breaks torch.compile and is much less performant. Remove SpaceToDepthModule

2024-05-06 16:32:49 -07:00

mlp.py

…

non_local_attn.py

…

norm_act.py

Add ConvStem and MobileCLIP hybrid model for B variant. Add full norm disable support to ConvNormAct layers

2024-06-06 09:15:27 -07:00

norm.py

Remove JIT activations, take jit out of ME activations. Remove other instances of torch.jit.script. Breaks torch.compile and is much less performant. Remove SpaceToDepthModule

2024-05-06 16:32:49 -07:00

padding.py

…

patch_dropout.py

Credit earlier work with the same idea.

2024-05-16 22:50:34 +02:00

patch_embed.py

…

pool2d_same.py

…

pos_embed_rel.py

…

pos_embed_sincos.py

Mapping OpenAI CLIP Modified ResNet weights -> ByobNet. Improve AttentionPool2d layers. Fix #1731

2024-06-09 16:54:48 -07:00

pos_embed.py

Adding pos embed resize fns to FX autowrap exceptions

2024-06-10 12:06:47 -07:00

selective_kernel.py

Remove separate ConvNormActAa class, merge with ConvNormAct

2024-06-10 12:05:35 -07:00

separable_conv.py

…

space_to_depth.py

Remove JIT activations, take jit out of ME activations. Remove other instances of torch.jit.script. Breaks torch.compile and is much less performant. Remove SpaceToDepthModule

2024-05-06 16:32:49 -07:00

split_attn.py

…

split_batchnorm.py

…

squeeze_excite.py

…

std_conv.py

…

test_time_pool.py

…

trace_utils.py

…

typing.py

…

weight_init.py

…