Commits
- 203acee: rename print_interval to log_interval (Zhiyuan Chen)
- a5d55f9: fix wraps in flexible_decorator (Zhiyuan Chen)
- c0fabfb: rename score_set to score_split (Zhiyuan Chen)
- 48ef85b: rename AverageMeters to MultiTaskAverageMeter (Zhiyuan Chen)
- b1b8a21: add MultiTaskMetrics (Zhiyuan Chen)
- 960f08c: do not save checkpoint if save_interval is None (Zhiyuan Chen)
- 2eaa06d: auto set log_interval if None (Zhiyuan Chen)
- 52692e8: fix dataloader is not auto prepared in some cases (Zhiyuan Chen)
- bb0655c: no longer print all results at the end of training (Zhiyuan Chen)
- 89d55f8: catch ValueError in r2 (Zhiyuan Chen)
- 8bc5731: support format_spec in format_result (Zhiyuan Chen)
- 800abd0: squeeze output before metrics.update (Zhiyuan Chen)
- 885e81b: catch ValueError in pearson, spearman, and mcc (Zhiyuan Chen)
- 41a43b0: support NestedTensor.concat (Zhiyuan Chen)
- 63e8ef6: add load_pandas (Zhiyuan Chen)
- 3dcd876: fix runner when dataloader missing in config (Zhiyuan Chen)
- df99b35: fix NestedTensor._storage use torch.Tensor in some cases (Zhiyuan Chen)
- 7624639: add dl.tensor API since PyTorch changed the constructor of Tensor (Zhiyuan Chen)
- 5754169: fix docs (Zhiyuan Chen)