Ross Wightman 962958723c More Hiera updates. Add forward_intermediates to hieradat/sam2 impl. Make both use same classifier module. Add coarse bool to intermediates.

2024-08-16 11:10:04 -07:00

..

__init__.py

Rename global pos embed for Hiera abswin, factor out commonly used vit weight init fns to layers. Add a channels-last ver of normmlp head.

2024-08-15 17:46:36 -07:00

activations_me.py

…

activations.py

…

adaptive_avgmax_pool.py

…

attention2d.py

Add xavier_uniform init of MNVC hybrid attention modules. Small improvement in training stability.

2024-07-26 17:03:40 -07:00

attention_pool2d.py

…

attention_pool.py

…

blur_pool.py

…

bottleneck_attn.py

…

cbam.py

…

classifier.py

More Hiera updates. Add forward_intermediates to hieradat/sam2 impl. Make both use same classifier module. Add coarse bool to intermediates.

2024-08-16 11:10:04 -07:00

cond_conv2d.py

…

config.py

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

conv2d_same.py

…

conv_bn_act.py

…

create_act.py

Add WIP HieraDet impl (SAM2 backbone support)

2024-08-15 17:58:15 -07:00

create_attn.py

…

create_conv2d.py

…

create_norm_act.py

…

create_norm.py

…

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

…

hybrid_embed.py

…

inplace_abn.py

…

interpolate.py

Add more advanced interpolation method from BEiT and support non-square window & image size adaptation for

2023-08-08 16:41:16 -07:00

lambda_layer.py

…

layer_scale.py

Fix hiera init with num_classes=0, fix weight tag names for sbb2 hiera/vit weights, add LayerScale/LayerScale2d to layers

2024-08-15 11:14:38 -07:00

linear.py

…

median_pool.py

…

mixed_conv2d.py

…

ml_decoder.py

…

mlp.py

…

non_local_attn.py

…

norm_act.py

…

norm.py

…

padding.py

…

patch_dropout.py

…

patch_embed.py

…

pool2d_same.py

…

pos_embed_rel.py

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

pos_embed_sincos.py

…

pos_embed.py

…

selective_kernel.py

…

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

Rename global pos embed for Hiera abswin, factor out commonly used vit weight init fns to layers. Add a channels-last ver of normmlp head.

2024-08-15 17:46:36 -07:00