Issues
- 6
🐛[BUG]: Some NCCL operations have failed or timed out. Due to the asynchronous nature of CUDA kernels, subsequent GPU operations might run on corrupted/incomplete data.
#691 opened by wlu1998 - 3
🐛[BUG]: aero_graph_net failed to load dataset
#710 opened by willyawan16 - 0
- 2
Questions about CorrDiff example
#706 opened by MyGitHub-G - 4
Question about the CorrDiff Example
#704 opened by MyGitHub-G - 1
🐛[BUG]: FigConvNet Reproduce
#705 opened by wangguan1995 - 0
🐛[BUG]: Instantiating `EDMPrecondSR` fails from `TypeError: got an unexpected keyword argument 'scale_cond_input'` with `modulus==0.8.0` (Modulus container 24.09)
#694 opened by luke-conibear - 0
🐛[BUG]: Importing opencv-python fails when using Modulus container (24.09)
#693 opened by luke-conibear - 6
- 2
- 4
🐛[BUG]: IndexError: list index out of range when training BiStride MeshGraphNet
#695 opened by AndreaPi - 0
🐛[BUG]: Transolver MLFlow function not exported
#699 opened by alexk101 - 3
🐛[BUG]: CorrDiff loss is scaled by hyper-parameter
#605 opened by chychen - 2
- 1
- 3
⛰️[EPIC]: CorrDiff Usability Enhancements
#589 opened by mnabian - 1
🚀[FEA]: Add transolver model.
#592 opened by luohk19 - 1
- 2
- 0
Corrdiff: cleanup CWB dataloader.
#599 opened by mnabian - 0
Unit tests for the CorrDiff dataloader
#601 opened by mnabian - 2
- 0
CorrDiff: Comprehensive Documentation
#632 opened by mnabian - 0
CorrDiff: Better handling of the downsampling factor
#634 opened by mnabian - 0
CorrDiff training and inference validation
#655 opened by mnabian - 0
- 6
🐛[BUG]: ERA5 dataset_download example fails
#631 opened by negedng - 4
- 1
🐛[BUG]: datapipes.healpix is not importable
#680 opened by yairchn - 1
⛰️[EPIC]: GenCast
#569 opened by mnabian - 1
Installation issue on windows wsl
#657 opened by AvisP - 0
🐛[BUG]: Add steps to install `shapely`
#596 opened by ktangsali - 0
CorrDiff: Implement a synthetic datapipe
#598 opened by mnabian - 0
CorrDiff: Don't make models fully configurable, but make different versions (UNet_S, UNet_XS, etc.)
#602 opened by mnabian - 1
CorrDiff: Reduce redundancy between generate.py, train.py and training_loop.py
#633 opened by mnabian - 1
CorrDiff: Split generate.py, train.py and training_loop.py into functions/classes that handle more well-defined tasks
#639 opened by mnabian - 0
CorrDiff-Lite recipe
#635 opened by mnabian - 2
- 0
Train GraphCast with the GraphTransformer Processor
#580 opened by mnabian - 1
- 0
- 1
CorrDiff: Use more descriptive parameter names
#600 opened by mnabian - 1
- 1
CorrDiff: Remove unused modules/code
#612 opened by mnabian - 1
CorrDiff: Switch to Tensorboard logging
#622 opened by mnabian - 1
CorrDiff: Remove unnecessary EDM abstractions
#613 opened by mnabian - 1
- 1
CorrDiff: Better handling of EMA
#623 opened by mnabian - 0
Implement a Graph Transformer model
#572 opened by mnabian - 0
CorrDiff: Improve config handling, separate configs for different dataloaders, reduce configurable parameters and use the defaults instead
#614 opened by mnabian