151679c2f1
Add custom grad tests, fix cut & paste error with hard_mish ME, add a few more pytorch act fns to factory
2020-06-11 14:49:23 -07:00
6c7932fe75
Update sotabench.py, tweak VovNet cfg
2020-06-11 14:34:25 -07:00
d3ee3de96a
Update validation script first batch prime and clear cuda cache between multi-model runs
2020-06-11 13:34:21 -07:00
0aca08384f
Update regnet cfg keys to match model names so registry works properly
2020-06-09 14:37:45 -07:00
b41b8d0108
Update results csv files
2020-06-09 14:36:24 -07:00
13c0a6290e
Add ESE-VovNet39 weights, 79.32 top-1
2020-06-06 22:50:21 -07:00
4ddde1d3a4
Fix two regressions
2020-06-05 11:04:51 -07:00
a7e8cadd15
Remove pointless densenet configs, add an iabn version of 264 as it makes more sense to try someday...
2020-06-03 17:13:52 -07:00
e78daf586a
better densenet121 and densenetblur121d weights
2020-06-03 13:30:03 -07:00
7be299504f
Add missing feature_info() on MobileNetV3, make hook feature output order/type consistent with bottleneck (list, decreasing fmap size)
2020-06-03 00:00:37 -07:00
88129b2569
Add set_layer_config contextmgr to adjust all layer configs at once, use in create_module with new args. Remove a few old warning causing constant annotations for jit.
2020-06-02 21:06:10 -07:00
f28170df3f
Fix an an untested change, remove a debug print
2020-06-01 17:26:42 -07:00
eb7653614f
Monster commit, activation refactor, VoVNet, norm_act improvements, more
...
* refactor activations into basic PyTorch, jit scripted, and memory efficient custom auto
* implement hard-mish, better grad for hard-swish
* add initial VovNet V1/V2 impl, fix #151
* VovNet and DenseNet first models to use NormAct layers (support BatchNormAct2d, EvoNorm, InplaceIABN)
* Wrap IABN for any models that use it
* make more models torchscript compatible (DPN, PNasNet, Res2Net, SelecSLS) and add tests
2020-06-01 17:16:52 -07:00
ff94ffce61
Merge branch 'master' into densenet_update_and_more
2020-05-26 00:02:57 -07:00
5966654052
Update requirements so PyTorch 1.4 is min, add separate sotabench reqs to force 1.4 and prevent error with 1.5
2020-05-24 13:28:03 -07:00
0ea53cecc3
Merge branch 'master' into densenet_update_and_more
2020-05-22 16:18:10 -07:00
6441e9cc1b
Fix memory_efficient mode for DenseNets. Add AntiAliasing (Blur) support for DenseNets and create one test model. Add lr cycle/mul params to train args.
2020-05-22 16:16:45 -07:00
d79ac48626
Update sotabench.py
2020-05-22 14:42:43 -07:00
e881383b4e
Merge pull request #154 from rwightman/tests_bugfixes
...
Add backward and default_cfg tests and fix a few issues found. Fix #153
2020-05-21 19:44:12 -07:00
4d13db538f
Add x48d ResNext101s to test exclude for ubuntu
2020-05-21 19:13:41 -07:00
4212cd3b9f
Another attempt at getting Ubuntu test runner to work
2020-05-21 18:55:10 -07:00
20329f2630
Bring down test resolutions to see if we can at least do a fwd on the L2 models
2020-05-21 16:49:46 -07:00
3873ea710e
Minor test change
2020-05-21 15:51:47 -07:00
afb6bd0669
Add backward and default_cfg tests and fix a few issues found. Fix #153
2020-05-21 15:30:58 -07:00
ea2e59ca36
Merge pull request #150 from rwightman/regnet
...
Add RegNet models and weights
2020-05-18 00:26:40 -07:00
50658b9a67
Add RegNet models and weights
2020-05-18 00:08:52 -07:00
7df83258c9
Merge branch 'master' into densenet_update_and_more
2020-05-13 23:34:44 -07:00
dab9935b36
Merge pull request #148 from rwightman/drop_block_improve
...
Improve dropblock impl, add fast variant, better AMP speed, inplace…
2020-05-13 16:30:34 -07:00
1904ed8fec
Improve dropblock impl, add fast variant, and better AMP speed, inplace, batchwise... few ResNeSt cleanups
2020-05-13 15:17:08 -07:00
63addb741f
Merge pull request #146 from rwightman/inceptionv3_fix
...
Remove annoying torchvision InceptionV3 dependency on scipy and insanely slow tru…
2020-05-12 22:03:37 -07:00
17270c69b9
Remove annoying InceptionV3 dependency on scipy and insanely slow trunc_norm init. Bring InceptionV3 code into this codebase and use upcoming torch trunch_norm_ init.
2020-05-12 21:59:34 -07:00
c4ca016656
Merge pull request #145 from rwightman/resnest
...
ResNeSt
2020-05-12 13:38:29 -07:00
208e7912f7
Missed one of the abalation model entrypoints, update README
2020-05-12 13:36:31 -07:00
5bd1ad13e7
Refactor test indent
2020-05-12 13:07:46 -07:00
9cc289f18c
Exclude EfficientNet-L2 models from test
2020-05-12 13:07:03 -07:00
2f884a0ce5
Add resnest14, resnest26, and two of the abalation grouped resnest50 models
2020-05-12 12:21:52 -07:00
f4cdc2ac31
Add ResNeSt models
2020-05-11 23:27:09 -07:00
780860d140
Add norm_act factory method, move JIT of norm layers to factory
2020-05-09 22:09:21 -07:00
14edacdf9a
DenseNet converted to support ABN (norm + act) modules. Experimenting with EvoNorm, IABN
2020-05-09 18:26:41 -07:00
e545bb9401
Update test_inference.py
...
Not six min
2020-05-07 10:15:49 -07:00
305a2db705
Update test_inference.py
...
Make the timeout 5-min for now, see if we can get a pass...
2020-05-07 10:14:24 -07:00
022ed001f3
Update DenseNet to latest in Torchvision (torchscript compat, checkpointing, proper init). Start adding ehanced configurability, stem options...
2020-05-07 09:57:48 -07:00
14e01b878c
Merge pull request #143 from michalwols/master
...
Setup Github Action to instantiate and run a forward pass with each registered model.
2020-05-07 09:47:13 -07:00
8c77f14cae
Install cpu version of torch on ubuntu
2020-05-07 01:09:16 -04:00
69d725c9fe
Basic forward pass test for all registered models
2020-05-07 00:20:58 -04:00
8da43e0617
Install extra dependencies required by some models and log test durations
2020-05-07 00:20:34 -04:00
16f1f77b41
Add a test workflow for github actions
2020-05-06 23:21:50 -04:00
a7ebe09029
Replace all None by nn.Identity() in all models reset_classifier when False-values num_classes is given.
...
Make small code refactoring
2020-05-06 09:54:03 +03:00
6cc11a8821
Merge pull request #141 from Animatory/fix_HRNet
...
Fixed HRNet modules
2020-05-05 18:21:34 -07:00
f0eb021620
Replace all None by nn.Identity() in HRNet modules
2020-05-05 21:09:35 +03:00
Ross Wightman
Ross Wightman
michal
Mike
Vyacheslav Shults