你好。我训练COCO数据集，发现先关键点loss居高不下

[mshmoon]

你好，我用您的代码训练COCOC数据集，发现'loss_classifier'会下降，但是 'loss_keypoint'基本居高不下

[scnuhealthy]

可以发下你的config和log吗，可能是lr的问题

[mshmoon]

您好，我发现COCO数据集收敛了。但是我自己用labelme标注了2张图片，其中有人的bounding box和3个keypoints 但是不收敛，好难过，方便吗，大佬，加一下qq

[mshmoon]

大佬，我QQ是370308707

[hasayake007]

我跟你正相反,loss几乎为0,但是评估的时候AP也为0

[yijiudd]

求问 我只修改了项目中数据集的路径，但是按照默认参数train结束后，ap结果很差，loss一直维持在5左右不变，是什么原因啊，大佬

Test: Total time: 0:05:23 (0.0648 s / it) Averaged stats: model_time: 0.0573 (0.0591) evaluator_time: 0.0023 (0.0038) Accumulating evaluation results... DONE (t=1.17s). Accumulating evaluation results... DONE (t=0.24s). IoU metric: bbox Average Precision (AP) @[ IoU=0.50:0.95 | area= all | maxDets=100 ] = 0.173 Average Precision (AP) @[ IoU=0.50 | area= all | maxDets=100 ] = 0.410 Average Precision (AP) @[ IoU=0.75 | area= all | maxDets=100 ] = 0.115 Average Precision (AP) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = 0.128 Average Precision (AP) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.204 Average Precision (AP) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.205 Average Recall (AR) @[ IoU=0.50:0.95 | area= all | maxDets= 1 ] = 0.087 Average Recall (AR) @[ IoU=0.50:0.95 | area= all | maxDets= 10 ] = 0.257 Average Recall (AR) @[ IoU=0.50:0.95 | area= all | maxDets=100 ] = 0.353 Average Recall (AR) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = 0.260 Average Recall (AR) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.394 Average Recall (AR) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.447 IoU metric: keypoints Average Precision (AP) @[ IoU=0.50:0.95 | area= all | maxDets= 20 ] = 0.174 Average Precision (AP) @[ IoU=0.50 | area= all | maxDets= 20 ] = 0.388 Average Precision (AP) @[ IoU=0.75 | area= all | maxDets= 20 ] = 0.129 Average Precision (AP) @[ IoU=0.50:0.95 | area=medium | maxDets= 20 ] = 0.196 Average Precision (AP) @[ IoU=0.50:0.95 | area= large | maxDets= 20 ] = 0.165 Average Recall (AR) @[ IoU=0.50:0.95 | area= all | maxDets= 20 ] = 0.304 Average Recall (AR) @[ IoU=0.50 | area= all | maxDets= 20 ] = 0.602 Average Recall (AR) @[ IoU=0.75 | area= all | maxDets= 20 ] = 0.264 Average Recall (AR) @[ IoU=0.50:0.95 | area=medium | maxDets= 20 ] = 0.292 Average Recall (AR) @[ IoU=0.50:0.95 | area= large | maxDets= 20 ] = 0.320

[yijiudd]

因为我使用一个rtx4070训练的，所以lr应该改成0.0025吗

[scnuhealthy]

检查data, 降低lr等

[yijiudd]

检查data, 降低lr等

嗯嗯，目前使用lr=0.0025 results这样是正常吗： IoU metric: bbox Average Precision (AP) @[ IoU=0.50:0.95 | area= all | maxDets=100 ] = 0.522 Average Precision (AP) @[ IoU=0.50 | area= all | maxDets=100 ] = 0.806 Average Precision (AP) @[ IoU=0.75 | area= all | maxDets=100 ] = 0.567 Average Precision (AP) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = 0.355 Average Precision (AP) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.605 Average Precision (AP) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.681 Average Recall (AR) @[ IoU=0.50:0.95 | area= all | maxDets= 1 ] = 0.182 Average Recall (AR) @[ IoU=0.50:0.95 | area= all | maxDets= 10 ] = 0.537 Average Recall (AR) @[ IoU=0.50:0.95 | area= all | maxDets=100 ] = 0.616 Average Recall (AR) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = 0.465 Average Recall (AR) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.685 Average Recall (AR) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.760 IoU metric: keypoints Average Precision (AP) @[ IoU=0.50:0.95 | area= all | maxDets= 20 ] = 0.611 Average Precision (AP) @[ IoU=0.50 | area= all | maxDets= 20 ] = 0.837 Average Precision (AP) @[ IoU=0.75 | area= all | maxDets= 20 ] = 0.663 Average Precision (AP) @[ IoU=0.50:0.95 | area=medium | maxDets= 20 ] = 0.560 Average Precision (AP) @[ IoU=0.50:0.95 | area= large | maxDets= 20 ] = 0.691 Average Recall (AR) @[ IoU=0.50:0.95 | area= all | maxDets= 20 ] = 0.681 Average Recall (AR) @[ IoU=0.50 | area= all | maxDets= 20 ] = 0.888 Average Recall (AR) @[ IoU=0.75 | area= all | maxDets= 20 ] = 0.731 Average Recall (AR) @[ IoU=0.50:0.95 | area=medium | maxDets= 20 ] = 0.631 Average Recall (AR) @[ IoU=0.50:0.95 | area= large | maxDets= 20 ] = 0.751

[yijiudd]

不过我使用pretrained_weights 训练，loss基本没有变，一直是3左右