Closed hi-zhangjie closed 1 year ago
Sorry, I didn't encounter this issue in this project.
But according to my experimence in mmdet3d, it seems a normal log and do not affect the training.
2022-09-24 15:05:12,648 - mmcv - INFO - Reducer buckets have been rebuilt in this iteration. 2022-09-24 15:05:28,102 - mmdet - INFO - Epoch [1][20/3953] lr: 1.000e-05, eta: 2 days, 2:41:27, time: 1.283, data_time: 0.420, memory: 15217, task0.loss_heatmap: 31.7599, task0.loss_bbox: 101.8863, loss: 133.6462, grad_norm: 1067.8555 2022-09-24 15:05:44,301 - mmdet - INFO - Epoch [1][40/3953] lr: 1.002e-05, eta: 1 day, 17:20:47, time: 0.810, data_time: 0.005, memory: 15217, task0.loss_heatmap: 21.7878, task0.loss_bbox: 81.5346, loss: 103.3225, grad_norm: 649.0363 2022-09-24 15:06:00,045 - mmdet - INFO - Epoch [1][60/3953]
how to fix this problem