loss=0

[AImagicAI]

您好，我在训练自己数据集时，loss=0。

04/10 14:42:49 - mmengine - INFO - Epoch(train) [1][ 50/11841] lr: 2.0000e-04 eta: 17 days, 5:39:40 time: 0.5031 data_time: 0.0056 memory: 9182 loss: 0.8814 loss_rpn_cls: 0.3083 loss_rpn_bbox: 0.0000 loss_cls: 0.5731 acc: 100.0000 loss_bbox: 0.0000 loss_mask: 0.0000 04/10 14:42:54 - mmengine - INFO - Epoch(train) [1][ 60/11841] lr: 2.0000e-04 eta: 17 days, 1:57:52 time: 0.4753 data_time: 0.0024 memory: 9182 loss: 0.2264 loss_rpn_cls: 0.1712 loss_rpn_bbox: 0.0000 loss_cls: 0.0552 acc: 100.0000 loss_bbox: 0.0000 loss_mask: 0.0000 04/10 14:42:58 - mmengine - INFO - Epoch(train) [1][ 70/11841] lr: 2.0000e-04 eta: 17 days, 0:31:02 time: 0.4774 data_time: 0.0024 memory: 9182 loss: 0.0565 loss_rpn_cls: 0.0565 loss_rpn_bbox: 0.0000 loss_cls: 0.0000 acc: 100.0000 loss_bbox: 0.0000 loss_mask: 0.0000 04/10 14:43:03 - mmengine - INFO - Epoch(train) [1][ 80/11841] lr: 2.0000e-04 eta: 16 days, 23:08:44 time: 0.4789 data_time: 0.0025 memory: 9182 loss: 0.0067 loss_rpn_cls: 0.0067 loss_rpn_bbox: 0.0000 loss_cls: 0.0000 acc: 100.0000 loss_bbox: 0.0000 loss_mask: 0.0000 04/10 14:43:08 - mmengine - INFO - Epoch(train) [1][ 90/11841] lr: 2.0000e-04 eta: 16 days, 21:24:56 time: 0.4792 data_time: 0.0026 memory: 9182 loss: 0.0006 loss_rpn_cls: 0.0006 loss_rpn_bbox: 0.0000 loss_cls: 0.0000 acc: 100.0000 loss_bbox: 0.0000 loss_mask: 0.0000 04/10 14:43:13 - mmengine - INFO - Epoch(train) [1][ 100/11841] lr: 2.0000e-04 eta: 16 days, 20:03:36 time: 0.4797 data_time: 0.0027 memory: 9182 loss: 0.0002 loss_rpn_cls: 0.0002 loss_rpn_bbox: 0.0000 loss_cls: 0.0000 acc: 100.0000 loss_bbox: 0.0000 loss_mask: 0.0000 04/10 14:43:18 - mmengine - INFO - Epoch(train) [1][ 110/11841] lr: 2.0000e-04 eta: 16 days, 18:54:51 time: 0.4797 data_time: 0.0027 memory: 9182 loss: 0.0002 loss_rpn_cls: 0.0002 loss_rpn_bbox: 0.0000 loss_cls: 0.0000 acc: 100.0000 loss_bbox: 0.0000 loss_mask: 0.0000 04/10 14:43:22 - mmengine - INFO - Epoch(train) [1][ 120/11841] lr: 2.0000e-04 eta: 16 days, 17:53:47 time: 0.4775 data_time: 0.0026 memory: 9182 loss: 0.0001 loss_rpn_cls: 0.0001 loss_rpn_bbox: 0.0000 loss_cls: 0.0000 acc: 100.0000 loss_bbox: 0.0000 loss_mask: 0.0000 04/10 14:43:27 - mmengine - INFO - Epoch(train) [1][ 130/11841] lr: 2.0000e-04 eta: 16 days, 17:06:29 time: 0.4761 data_time: 0.0026 memory: 9182 loss: 0.0001 loss_rpn_cls: 0.0001 loss_rpn_bbox: 0.0000 loss_cls: 0.0000 acc: 100.0000 loss_bbox: 0.0000 loss_mask: 0.0000 04/10 14:43:32 - mmengine - INFO - Epoch(train) [1][ 140/11841] lr: 2.0000e-04 eta: 16 days, 16:18:34 time: 0.4757 data_time: 0.0026 memory: 9182 loss: 0.0001 loss_rpn_cls: 0.0001 loss_rpn_bbox: 0.0000 loss_cls: 0.0000 acc: 100.0000 loss_bbox: 0.0000 loss_mask: 0.0000 04/10 14:43:37 - mmengine - INFO - Epoch(train) [1][ 150/11841] lr: 2.0000e-04 eta: 16 days, 15:42:11 time: 0.4755 data_time: 0.0026 memory: 9182 loss: 0.0001 loss_rpn_cls: 0.0001 loss_rpn_bbox: 0.0000 loss_cls: 0.0000 acc: 100.0000 loss_bbox: 0.0000 loss_mask: 0.0000 04/10 14:43:41 - mmengine - INFO - Epoch(train) [1][ 160/11841] lr: 2.0000e-04 eta: 16 days, 15:11:23 time: 0.4755 data_time: 0.0026 memory: 9182 loss: 0.0001 loss_rpn_cls: 0.0001 loss_rpn_bbox: 0.0000 loss_cls: 0.0000 acc: 100.0000 loss_bbox: 0.0000 loss_mask: 0.0000

[lphilomena]

你好！我也遇到了同样的问题，请问你解决这个问题了吗？

[KyanChen]

fixed: 请查看QA中的第6条。见 https://github.com/KyanChen/RSPrompter/blob/release/README_zh-CN.md