c3fbdd4655
Fix efficient head for MobileNetV3
6 years ago
17da1adaca
A few MobileNetV3 tweaks
...
* fix expansion ratio on early block
* change comment re stride mistake in paper
* fix rounding not being called properly for all multipliers != 1.0
6 years ago
6523e4abe4
More appropriate name for se channel
6 years ago
db056d97e2
Add MobileNetV3 and associated changes hard-swish, hard-sigmoid, efficient head, etc
6 years ago
02abeb95bf
Add the url for tflite mnasnet ported weights
6 years ago
0956cd4b66
Update README.md
...
Add notes for latest mobile model weights...
6 years ago
4663fc2132
Add support for tflite mnasnet pretrained weights and included spnasnet pretrained weights of my own.
...
* tensorflow 'SAME' padding support added to GenMobileNet models for tflite pretrained weights
* folded batch norm support (made batch norm optional and enable conv bias) for tflite pretrained weights
* add url for spnasnet1_00 weights that I recently trained
* fix SE reduction size for semnasnet models
6 years ago
afb357ff68
Make genmobilenet weight init switchable, fix fan_out in google style linear init
6 years ago
0a853990e7
Add distributed sampler that maintains order of original dataset (for validation)
6 years ago
8fbd62a169
Exclude batchnorm and bias params from weight_decay by default
6 years ago
34cd76899f
Add Single-Path NAS pixel1 model
6 years ago
419555be62
Update a few GenMobileNet comments
6 years ago
1cf3ea0467
Update README.md
6 years ago
bc264269c9
Morph mnasnet impl into a generic mobilenet that covers Mnasnet, MobileNetV1/V2, ChamNet, FBNet, and related
...
* add an alternate RMSprop opt that applies eps like TF
* add bn params for passing through alternates and changing defaults to TF style
6 years ago
e9c7961efc
Fix pooling in mnasnet, more sensible default for AMP opt level
6 years ago
996c77aa94
Prep mnasnet for pretrained models, use the select global pool, some comment mistakes
6 years ago
6b4f9ba223
Add MNASNet A1, B1, and Small models as per the TF impl. Testing/training in progress...
6 years ago
c88e80081d
Fix missing cfg key check
6 years ago
073d31a076
Update README.md
6 years ago
7ba78aaaeb
Update README.md
6 years ago
e8e8bce335
Create README.md
6 years ago
9e296dbffb
Add seresnet26_32x4d cfg and weights + interpolation str->PIL enum fn
6 years ago
71afec86d3
Loader tweaks
6 years ago
79f615639e
Add pretrained weights for seresnet18
6 years ago
8a33a6c90a
Add checkpoint clean script, add link to pretrained resnext50 weights
6 years ago
6e9697eb9c
Fix small bug in seresnet input size and eval transform handling of img size
6 years ago
db1fe34d0c
Update a few comment, add some references
6 years ago
0562b91c38
Add per model crop pct, interpolation defaults, tie it all together
...
* create one resolve fn to pull together model defaults + cmd line args
* update attribution comments in some models
* test update train/validation/inference scripts
6 years ago
c328b155e9
Random erasing crash fix and args pass through
6 years ago
9c3859fb9c
Uniform pretrained model handling.
...
* All models have 'default_cfgs' dict
* load/resume/pretrained helpers factored out
* pretrained load operates on state_dict based on default_cfg
* test all models in validate
* schedule, optim factor factored out
* test time pool wrapper applied based on default_cfg
6 years ago
63e677d03b
Merge branch 'master' of github.com:rwightman/pytorch-models
6 years ago
0bc50e84f8
Lots of refactoring and cleanup.
...
* Move 'test time pool' to Module that can be used by any model, remove from DPN
* Remove ResNext model file and combine with ResNet
* Remove fbresnet200 as it was an old conversion and pretrained performance not worth param count
* Cleanup adaptive avgmax pooling and add back conctat variant
* Factor out checkpoint load fn
6 years ago
f1cd1a5ce3
Cleanup CheckpointSaver, add support for increasing or decreasing metric, switch to prec1 metric in train loop
6 years ago
c0e6e5f3db
Add common model interface to pnasnet and xception, update factory
6 years ago
f2029dfb65
Add smooth loss
6 years ago
b0158a593e
Fix distributed train script
6 years ago
183d8e4aef
Xception model working
6 years ago
1e23727f2f
Update inference script for new loader style
6 years ago
58571e992e
Change block avgpool in senets to mean for performance issues with NVIDIA and AMP especially
6 years ago
5180f94c7e
Distributed (multi-process) train, multi-gpu single process train, and NVIDIA AMP support
6 years ago
6f9a0c8ef2
Merge branch 'master' of github.com:rwightman/pytorch-models
6 years ago
5cb1a35c6b
Fixup Resnext, remove alternate shortcut types
6 years ago
d87824bd65
Merge branch 'master' of github.com:rwightman/pytorch-models
6 years ago
45cde6f0c7
Improve creation of data pipeline with prefetch enabled vs disabled, fixup inception_res_v2 and dpn models
6 years ago
321435e6b4
Update resnext init
6 years ago
2295cf56c2
Add some Nvidia performance enhancements (prefetch loader, fast collate), and refactor some of training and model fact/transforms
6 years ago
9d927a389a
Add adabound, random erasing
6 years ago
1577c52976
Resnext added, changes to bring it and seresnet in line with rest of models
6 years ago
e0cfeb7d8e
Add some models, remove a model, tweak some models
6 years ago
31055466fc
Fixup validate/inference script args, fix senet init for better test accuracy
6 years ago
Ross Wightman
Ross Wightman