ignite
ignite copied to clipboard
pytorch
High-level library to help with training and evaluating neural networks in PyTorch flexibly and transparently.
Description: To resolve warnings : `np` is deprecated Check list: - [ ] New tests are added (if a new feature is added) - [ ] New doc strings: description...
## 🐛 Bug description [Because of the division here](https://github.com/pytorch/ignite/blob/master/ignite/engine/__init__.py#L185), and similar functions, the loss has the wrong scale when using `gradient_accumulation_steps`. This makes it confusingly low in comparison to the...
Fixes #2703 Description: Editing FastaiLRFinder to have more than one parameter
Description: Modify to all_gather takes `group` Check list: - [ ] New tests are added (if a new feature is added) - [ ] New doc strings: description and/or example...
Description: Modify all_reduce to take `group` Check list: - [ ] New tests are added (if a new feature is added) - [ ] New doc strings: description and/or example...
```python optimizer = optim.SGD([ {'params': model.conv.parameters(), 'lr': 1}, {'params': model.linear.parameters(), 'lr': 0.1}, ], lr=3e-4, momentum=0.9) ```` Such as this, the optimizer has two different groups. Can anyone give an example?
Description: - Fixed removal nvidia-ml.list Check list: - [ ] New tests are added (if a new feature is added) - [ ] New doc strings: description and/or example code...
Flaky core dump errors: - tests/ignite/metrics/test_recall.py::test_distrib_hvd - tests/ignite/metrics/test_ssim.py::test_distrib_hvd - tests/ignite/metrics/test_precision.py::test_distrib_hvd
Current Engine's behaviour when we resume the run from terminated state is the following: ```python from ignite.engine import Engine, Events from ignite.utils import setup_logger, logging def func(engine, batch): print(engine.state.epoch, engine.state.iteration,...
As claimed in the title, how can i implement multiple updates of discriminator before a single update of generator in the train_step(engine, data)? As a common strategy of training GAN,...
