Open tonyleala opened 3 years ago
BOOTSTRAP_DATASETS: [] BOOTSTRAP_MODEL: DEVICE: cuda WEIGHTS: '' CUDNN_BENCHMARK: false DATALOADER: ASPECT_RATIO_GROUPING: true FILTER_EMPTY_ANNOTATIONS: true NUM_WORKERS: 4 REPEAT_THRESHOLD: 0.0 SAMPLER_TRAIN: TrainingSampler DATASETS: CATEGORY_MAPS: {} CLASS_TO_MESH_NAME_MAPPING: {} PRECOMPUTED_PROPOSAL_TOPK_TEST: 1000 PRECOMPUTED_PROPOSAL_TOPK_TRAIN: 2000 PROPOSAL_FILES_TEST: [] PROPOSAL_FILES_TRAIN: [] TEST:
Hi @tonyleala, you might need to adjust the training schedule if you train on 4 GPUs. Please consider adjusting BASE_LR
and WARMUP_FACTOR
.
If you do not know the root cause of the problem, please post according to this template:
Instructions To Reproduce the Issue:
Check https://stackoverflow.com/help/minimal-reproducible-example for how to ask good questions. Simplify the steps to reproduce the issue using suggestions from the above link, and provide them below:
Expected behavior:
If there are no obvious crash in "full logs" provided above, please tell us the expected behavior.
If you expect a model to converge / work better, we do not help with such issues, unless a model fails to reproduce the results in detectron2 model zoo, or proves existence of bugs.
Environment:
Paste the output of the following command:
When I train the model just with your provided config, we may reach the point:
In addition, in the process of training, the loss densepose UV is sometimes less than zero.