Releases: w86763777/pytorch-image-generation-metrics
Releases · w86763777/pytorch-image-generation-metrics
Improve progress bar for notebook
- use
tqdm.auto
instead oftqdm
- use argument name instead of positional parameter for
calc_fid_stats
- update
version
Compatiable with torchvision==0.13.1
- Support
torchvision==0.13.1
. - Improve documents.
- Make unittest self contained.
- Add wheel to release
Compatiable with torchvision==0.11.1
- Fix type hint
- Fix import error of
load_state_dict_from_url
- Add
--use_torch
to cli toolpytorch_gan_metrics/calc_metrics.py
- Update reproducing table in readme
Bug fix
Bug fix
- Fix CUDA is reinitialized in subprocess when multiprocess is used for distributed training.
- Add argument
device
with default valuetorch.device('cuda:0')
tocore.get_inception_feature
which can be assigned by**kwargs
for all utility functionsutils.*
.