f2e14685a8
Add force-cpu flag for train/validate, fix CPU fallback for device init, remove old force cpu flag for EMA model weights
3 years ago
0d82876132
Add comment for reference re PyTorch XLA 'race' issue
3 years ago
40457e5691
Transforms, augmentation work for bits, add RandomErasing support for XLA (pushing into transforms), revamp of transform/preproc config, etc ongoing...
3 years ago
c3db5f5801
Worker hack for TFDS eval, add TPU env var setting.
3 years ago
f411724de4
Fix checkpoint delete issue. Add README about bits and initial Pytorch XLA usage on TPU-VM. Add some FIXMEs and fold train_cfg into train_state by default.
3 years ago
91ab0b6ce5
Add proper TrainState checkpoint save/load. Some reorg/refactoring and other cleanup. More to go...
3 years ago
5b9c69e80a
Add basic training resume based on legacy code
3 years ago
cbd4ee737f
Fix model init for XLA, remove some prints.
3 years ago
6d90fcf282
Fix distribute_bn and model_ema
3 years ago
aa92d7b1c5
Major timm.bits update. Updater and DeviceEnv now dataclasses, after_step closure used, metrics base impl w/ distributed reduce, many tweaks/fixes.
3 years ago
76de984a5f
Fix some bugs with XLA support, logger, add hacky xla dist launch script since torch.dist.launch doesn't work
3 years ago
12d9a6d4d2
First timm.bits commit, add initial abstractions, WIP updates to train, val... some of it working
3 years ago
e685618f45
Merge pull request #550 from amaarora/wandb
...
Wandb Support
3 years ago
7c97e66f7c
Remove commented code, add more consistent seed fn
3 years ago
5772c55c57
Make wandb optional
3 years ago
f54897cc0b
make wandb not required but rather optional as huggingface_hub
3 years ago
f13f7508a9
Keep changes to minimal and use args.experiment as wandb project name if it exists
3 years ago
f8bb13f640
Default project name to None
3 years ago
3f028ebc0f
import wandb in summary.py
3 years ago
a9e5d9e5ad
log loss as before
3 years ago
624c9b6949
log to wandb only if using using wandb
3 years ago
00c8e0b8bd
Make use of wandb configurable
3 years ago
8e6fb861e4
Add wandb support
3 years ago
37c71a5609
Some further create_optimizer_v2 tweaks, remove some redudnant code, add back safe model str. Benchmark step times per batch.
3 years ago
288682796f
Update benchmark script to add precision arg. Fix some downstream (DeiT) compat issues with latest changes. Bump version to 0.4.7
3 years ago
a5310a3451
Merge remote-tracking branch 'origin/benchmark-fixes-vit_hybrids' into pit_and_vit_update
3 years ago
e2e3290fbf
Add '--experiment' to train args for fixed exp name if desired, 'train' not added to output folder if specified.
3 years ago
d584e7f617
Support for huggingface hub via create_model and default_cfgs.
...
* improve consistency of model creation helper fns
* add comments to some of the model helpers
* support passing external default_cfgs so they can be sourced from hub
3 years ago
2db2d87ff7
Add epoch-repeats arg to multiply the number of dataset passes per epoch. Currently for iterable datasets (read TFDS wrapper) only.
3 years ago
0e16d4e9fb
Add benchmark.py script, and update optimizer factory to be more friendly to use outside of argparse interface.
3 years ago
01653db104
Missed clip-mode arg for repo train script
3 years ago
4f49b94311
Initial AGC impl. Still testing.
3 years ago
d8e69206be
Merge pull request #419 from rwightman/byob_vgg_models
...
More models, GPU-Efficient Nets, RepVGG, classic VGG, and flexible Byob backbone.
3 years ago
0356e773f5
Default to native PyTorch AMP instead of APEX amp. Too many APEX issues cropping up lately.
3 years ago
5114c214fc
Change the Python interpreter to Python 3.x in the scripts
3 years ago
4203efa36d
Fix #387 so that checkpoint saver works with max history of 1. Add checkpoint-hist arg to train.py.
3 years ago
38d8f67570
Fix potential issue with change to num_classes arg in train/validate.py defaulting to None (rely on model def / default_cfg)
3 years ago
5d4c3d0af3
Add enhanced ParserImageInTar that can read images from tars within tars, folders with multiple tars, etc. Additional comment cleanup.
3 years ago
9d5d4b8df6
Fix silly train.py typo during dataset work
3 years ago
855d6cc217
More dataset work including factories and a tensorflow datasets (TFDS) wrapper
...
* Add parser/dataset factory methods for more flexible dataset & parser creation
* Add dataset parser that wraps TFDS image classification datasets
* Tweak num_classes handling bug for 21k models
* Add initial deit models so they can be benchmarked in next csv results runs
3 years ago
fd9061dbf7
Remove debug print from train.py
3 years ago
59ec7e6a53
Merge branch 'master' into imagenet21k_datasets_more
3 years ago
e42b140ade
Add --input-size option to scripts to specify full input dimensions from command-line
3 years ago
231d04e91a
ResNetV2 pre-act and non-preact model, w/ BiT pretrained weights and support for ViT R50 model. Tweaks for in21k num_classes passing. More to do... tests failing.
3 years ago
de6046e213
Initial commit for dataset / parser reorg to support additional datasets / types
3 years ago
2ed8f24715
A few more changes for 0.3.2 maint release. Linear layer change for mobilenetv3 and inception_v3, support no bias for linear wrapper.
4 years ago
460eba7f24
Work around casting issue with combination of native torch AMP and torchscript for Linear layers
4 years ago
27bbc70d71
Add back old ModelEma and rename new one to ModelEmaV2 to avoid compat breaks in dependant code. Shuffle train script, add a few comments, remove DataParallel support, support experimental torchscript training.
4 years ago
9214ca0716
Simplifying EMA...
4 years ago
80078c47bb
Add Adafactor and Adahessian optimizers, cleanup optimizer arg passing, add gradient clipping support.
4 years ago
Ross Wightman
Ross Wightman
Aman Arora
Csaba Kertesz