yolov8 训练时cpu占用过低，120核的cpu只有两核在满载跑，通过修改train_num_works的数量也无效

Prerequisite

[X] I have searched the existing and past issues but cannot get the expected help.
[X] I have read the FAQ documentation but cannot get the expected help.
[X] The bug has not been fixed in the latest version.

🐞 Describe the bug

Environment

_file_client_args = dict(backend='disk') _multiscale_resize_transforms = [ dict( transforms=[ dict(scale=( 640, 640, ), type='YOLOv5KeepRatioResize'), dict( allow_scale_up=False, pad_val=dict(img=114), scale=( 640, 640, ), type='LetterResize'), ], type='Compose'), dict( transforms=[ dict(scale=( 320, 320, ), type='YOLOv5KeepRatioResize'), dict( allow_scale_up=False, pad_val=dict(img=114), scale=( 320, 320, ), type='LetterResize'), ], type='Compose'), dict( transforms=[ dict(scale=( 960, 960, ), type='YOLOv5KeepRatioResize'), dict( allow_scale_up=False, pad_val=dict(img=114), scale=( 960, 960, ), type='LetterResize'), ], type='Compose'), ] affine_scale = 0.5 albu_train_transforms = [ dict(p=0.01, type='Blur'), dict(p=0.01, type='MedianBlur'), dict(p=0.01, type='ToGray'), dict(p=0.01, type='CLAHE'), ] base_lr = 0.01 batch_shapes_cfg = None class_name = ('coal_machine', ) close_mosaic_epochs = 5 custom_hooks = [ dict( ema_type='ExpMomentumEMA', momentum=0.0001, priority=49, strict_load=False, type='EMAHook', update_buffers=True), dict( switch_epoch=39995, switch_pipeline=[ dict( file_client_args=dict(backend='disk'), type='LoadImageFromFile'), dict(type='LoadAnnotations', with_bbox=True), dict(scale=( 640, 640, ), type='YOLOv5KeepRatioResize'), dict( allow_scale_up=True, pad_val=dict(img=114.0), scale=( 640, 640, ), type='LetterResize'), dict( border_val=( 114, 114, 114, ), max_aspect_ratio=100, max_rotate_degree=0.0, max_shear_degree=0.0, scaling_ratio_range=( 0.5, 1.5, ), type='YOLOv5RandomAffine'), dict( bbox_params=dict( format='pascal_voc', label_fields=[ 'gt_bboxes_labels', 'gt_ignore_flags', ], type='BboxParams'), keymap=dict(gt_bboxes='bboxes', img='image'), transforms=[ dict(p=0.01, type='Blur'), dict(p=0.01, type='MedianBlur'), dict(p=0.01, type='ToGray'), dict(p=0.01, type='CLAHE'), ], type='mmdet.Albu'), dict(type='YOLOv5HSVRandomAug'), dict(prob=0.5, type='mmdet.RandomFlip'), dict( meta_keys=( 'img_id', 'img_path', 'ori_shape', 'img_shape', 'flip', 'flip_direction', ), type='mmdet.PackDetInputs'), ], type='mmdet.PipelineSwitchHook'), ] data_root = './data/coal_machine_20230817/' dataset_type = 'YOLOv5CocoDataset' deepen_factor = 0.33 default_hooks = dict( checkpoint=dict( interval=10, max_keep_ckpts=2, save_best='auto', type='CheckpointHook'), logger=dict(interval=5, type='LoggerHook'), param_scheduler=dict( lr_factor=0.01, max_epochs=40000, scheduler_type='linear', type='YOLOv5ParamSchedulerHook', warmup_mim_iter=10), sampler_seed=dict(type='DistSamplerSeedHook'), timer=dict(type='IterTimerHook'), visualization=dict(type='mmdet.DetVisualizationHook')) default_scope = 'mmyolo' env_cfg = dict( cudnn_benchmark=True, dist_cfg=dict(backend='nccl'), mp_cfg=dict(mp_start_method='fork', opencv_num_threads=0)) file_client_args = dict(backend='disk') img_scale = ( 640, 640, ) img_scales = [ ( 640, 640, ), ( 320, 320, ), ( 960, 960, ), ] last_stage_out_channels = 1024 last_transform = [ dict( bbox_params=dict( format='pascal_voc', label_fields=[ 'gt_bboxes_labels', 'gt_ignore_flags', ], type='BboxParams'), keymap=dict(gt_bboxes='bboxes', img='image'), transforms=[ dict(p=0.01, type='Blur'), dict(p=0.01, type='MedianBlur'), dict(p=0.01, type='ToGray'), dict(p=0.01, type='CLAHE'), ], type='mmdet.Albu'), dict(type='YOLOv5HSVRandomAug'), dict(prob=0.5, type='mmdet.RandomFlip'), dict( meta_keys=( 'img_id', 'img_path', 'ori_shape', 'img_shape', 'flip', 'flip_direction', ), type='mmdet.PackDetInputs'), ] launcher = 'none' load_from = None log_level = 'INFO' log_processor = dict(by_epoch=True, type='LogProcessor', window_size=50) loss_bbox_weight = 7.5 loss_cls_weight = 0.5 loss_dfl_weight = 0.375 lr_factor = 0.01 max_aspect_ratio = 100 max_epochs = 40000 max_keep_ckpts = 2 metainfo = dict( classes=('coal_machine', ), palette=[ ( 20, 220, 60, ), ]) model = dict( backbone=dict( act_cfg=dict(inplace=True, type='SiLU'), arch='P5', deepen_factor=0.33, frozen_stages=4, last_stage_out_channels=1024, norm_cfg=dict(eps=0.001, momentum=0.03, type='BN'), type='YOLOv8CSPDarknet', widen_factor=0.5), bbox_head=dict( bbox_coder=dict(type='DistancePointBBoxCoder'), head_module=dict( act_cfg=dict(inplace=True, type='SiLU'), featmap_strides=[ 8, 16, 32, ], in_channels=[ 256, 512, 1024, ], norm_cfg=dict(eps=0.001, momentum=0.03, type='BN'), num_classes=1, reg_max=16, type='YOLOv8HeadModule', widen_factor=0.5), loss_bbox=dict( bbox_format='xyxy', iou_mode='ciou', loss_weight=7.5, reduction='sum', return_iou=False, type='IoULoss'), loss_cls=dict( loss_weight=0.5, reduction='none', type='mmdet.CrossEntropyLoss', use_sigmoid=True), loss_dfl=dict( loss_weight=0.375, reduction='mean', type='mmdet.DistributionFocalLoss'), prior_generator=dict( offset=0.5, strides=[ 8, 16, 32, ], type='mmdet.MlvlPointGenerator'), type='YOLOv8Head'), data_preprocessor=dict( bgr_to_rgb=True, mean=[ 0.0, 0.0, 0.0, ], std=[ 255.0, 255.0, 255.0, ], type='YOLOv5DetDataPreprocessor'), neck=dict( act_cfg=dict(inplace=True, type='SiLU'), deepen_factor=0.33, in_channels=[ 256, 512, 1024, ], norm_cfg=dict(eps=0.001, momentum=0.03, type='BN'), num_csp_blocks=3, out_channels=[ 256, 512, 1024, ], type='YOLOv8PAFPN', widen_factor=0.5), test_cfg=dict( max_per_img=300, multi_label=True, nms=dict(iou_threshold=0.7, type='nms'), nms_pre=30000, score_thr=0.001), train_cfg=dict( assigner=dict( alpha=0.5, beta=6.0, eps=1e-09, num_classes=1, topk=10, type='BatchTaskAlignedAssigner', use_ciou=True)), type='YOLODetector') model_test_cfg = dict( max_per_img=300, multi_label=True, nms=dict(iou_threshold=0.7, type='nms'), nms_pre=30000, score_thr=0.001) norm_cfg = dict(eps=0.001, momentum=0.03, type='BN') num_classes = 1 num_det_layers = 3 optim_wrapper = dict( clip_grad=dict(max_norm=10.0), constructor='YOLOv5OptimizerConstructor', loss_scale='dynamic', optimizer=dict( batch_size_per_gpu=8, lr=0.01, momentum=0.937, nesterov=True, type='SGD', weight_decay=0.0005), type='AmpOptimWrapper') param_scheduler = None persistent_workers = True pre_transform = [ dict(file_client_args=dict(backend='disk'), type='LoadImageFromFile'), dict(type='LoadAnnotations', with_bbox=True), ] resume = True save_epoch_intervals = 10 strides = [ 8, 16, 32, ] tal_alpha = 0.5 tal_beta = 6.0 tal_topk = 10 test_cfg = dict(type='TestLoop') test_dataloader = dict( batch_size=1, dataset=dict( ann_file='annotations/test.json', batch_shapes_cfg=None, data_prefix=dict(img='images/'), data_root='./data/coal_machine_20230817/', metainfo=dict(classes=('coal_machine', ), palette=[ ( 20, 220, 60, ), ]), pipeline=[ dict( file_client_args=dict(backend='disk'), type='LoadImageFromFile'), dict(scale=( 640, 640, ), type='YOLOv5KeepRatioResize'), dict( allow_scale_up=False, pad_val=dict(img=114), scale=( 640, 640, ), type='LetterResize'), dict(scope='mmdet', type='LoadAnnotations', with_bbox=True), dict( meta_keys=( 'img_id', 'img_path', 'ori_shape', 'img_shape', 'scale_factor', 'pad_param', ), type='mmdet.PackDetInputs'), ], test_mode=True, type='YOLOv5CocoDataset'), drop_last=False, num_workers=2, persistent_workers=True, pin_memory=True, sampler=dict(shuffle=False, type='DefaultSampler')) test_evaluator = dict( ann_file='./data/coal_machine_20230817/annotations/test.json', metric='bbox', proposal_nums=( 100, 1, 10, ), type='mmdet.CocoMetric') test_pipeline = [ dict(file_client_args=dict(backend='disk'), type='LoadImageFromFile'), dict(scale=( 640, 640, ), type='YOLOv5KeepRatioResize'), dict( allow_scale_up=False, pad_val=dict(img=114), scale=( 640, 640, ), type='LetterResize'), dict(scope='mmdet', type='LoadAnnotations', with_bbox=True), dict( meta_keys=( 'img_id', 'img_path', 'ori_shape', 'img_shape', 'scale_factor', 'pad_param', ), type='mmdet.PackDetInputs'), ] train_ann_file = 'annotations/instances_train2017.json' train_batch_size_per_gpu = 8 train_cfg = dict( dynamic_intervals=[ ( 490, 1, ), ], max_epochs=40000, type='EpochBasedTrainLoop', val_interval=5) train_data_prefix = 'train2017/' train_dataloader = dict( batch_size=8, collate_fn=dict(type='yolov5_collate'), dataset=dict( ann_file='annotations/trainval.json', data_prefix=dict(img='images/'), data_root='./data/coal_machine_20230817/', filter_cfg=dict(filter_empty_gt=False, min_size=32), metainfo=dict(classes=('coal_machine', ), palette=[ ( 20, 220, 60, ), ]), pipeline=[ dict( file_client_args=dict(backend='disk'), type='LoadImageFromFile'), dict(type='LoadAnnotations', with_bbox=True), dict( img_scale=( 640, 640, ), pad_val=114.0, pre_transform=[ dict( file_client_args=dict(backend='disk'), type='LoadImageFromFile'), dict(type='LoadAnnotations', with_bbox=True), ], type='Mosaic'), dict( border=( -320, -320, ), border_val=( 114, 114, 114, ), max_aspect_ratio=100, max_rotate_degree=0.0, max_shear_degree=0.0, scaling_ratio_range=( 0.5, 1.5, ), type='YOLOv5RandomAffine'), dict( bbox_params=dict( format='pascal_voc', label_fields=[ 'gt_bboxes_labels', 'gt_ignore_flags', ], type='BboxParams'), keymap=dict(gt_bboxes='bboxes', img='image'), transforms=[ dict(p=0.01, type='Blur'), dict(p=0.01, type='MedianBlur'), dict(p=0.01, type='ToGray'), dict(p=0.01, type='CLAHE'), ], type='mmdet.Albu'), dict(type='YOLOv5HSVRandomAug'), dict(prob=0.5, type='mmdet.RandomFlip'), dict( meta_keys=( 'img_id', 'img_path', 'ori_shape', 'img_shape', 'flip', 'flip_direction', ), type='mmdet.PackDetInputs'), ], type='YOLOv5CocoDataset'), num_workers=4, persistent_workers=True, pin_memory=True, sampler=dict(shuffle=True, type='DefaultSampler')) train_num_workers = 4 train_pipeline = [ dict(file_client_args=dict(backend='disk'), type='LoadImageFromFile'), dict(type='LoadAnnotations', with_bbox=True), dict( img_scale=( 640, 640, ), pad_val=114.0, pre_transform=[ dict( file_client_args=dict(backend='disk'), type='LoadImageFromFile'), dict(type='LoadAnnotations', with_bbox=True), ], type='Mosaic'), dict( border=( -320, -320, ), border_val=( 114, 114, 114, ), max_aspect_ratio=100, max_rotate_degree=0.0, max_shear_degree=0.0, scaling_ratio_range=( 0.5, 1.5, ), type='YOLOv5RandomAffine'), dict( bbox_params=dict( format='pascal_voc', label_fields=[ 'gt_bboxes_labels', 'gt_ignore_flags', ], type='BboxParams'), keymap=dict(gt_bboxes='bboxes', img='image'), transforms=[ dict(p=0.01, type='Blur'), dict(p=0.01, type='MedianBlur'), dict(p=0.01, type='ToGray'), dict(p=0.01, type='CLAHE'), ], type='mmdet.Albu'), dict(type='YOLOv5HSVRandomAug'), dict(prob=0.5, type='mmdet.RandomFlip'), dict( meta_keys=( 'img_id', 'img_path', 'ori_shape', 'img_shape', 'flip', 'flip_direction', ), type='mmdet.PackDetInputs'), ] train_pipeline_stage2 = [ dict(file_client_args=dict(backend='disk'), type='LoadImageFromFile'), dict(type='LoadAnnotations', with_bbox=True), dict(scale=( 640, 640, ), type='YOLOv5KeepRatioResize'), dict( allow_scale_up=True, pad_val=dict(img=114.0), scale=( 640, 640, ), type='LetterResize'), dict( border_val=( 114, 114, 114, ), max_aspect_ratio=100, max_rotate_degree=0.0, max_shear_degree=0.0, scaling_ratio_range=( 0.5, 1.5, ), type='YOLOv5RandomAffine'), dict( bbox_params=dict( format='pascal_voc', label_fields=[ 'gt_bboxes_labels', 'gt_ignore_flags', ], type='BboxParams'), keymap=dict(gt_bboxes='bboxes', img='image'), transforms=[ dict(p=0.01, type='Blur'), dict(p=0.01, type='MedianBlur'), dict(p=0.01, type='ToGray'), dict(p=0.01, type='CLAHE'), ], type='mmdet.Albu'), dict(type='YOLOv5HSVRandomAug'), dict(prob=0.5, type='mmdet.RandomFlip'), dict( meta_keys=( 'img_id', 'img_path', 'ori_shape', 'img_shape', 'flip', 'flip_direction', ), type='mmdet.PackDetInputs'), ] tta_model = dict( tta_cfg=dict(max_per_img=300, nms=dict(iou_threshold=0.65, type='nms')), type='mmdet.DetTTAModel') tta_pipeline = [ dict(file_client_args=dict(backend='disk'), type='LoadImageFromFile'), dict( transforms=[ [ dict( transforms=[ dict(scale=( 640, 640, ), type='YOLOv5KeepRatioResize'), dict( allow_scale_up=False, pad_val=dict(img=114), scale=( 640, 640, ), type='LetterResize'), ], type='Compose'), dict( transforms=[ dict(scale=( 320, 320, ), type='YOLOv5KeepRatioResize'), dict( allow_scale_up=False, pad_val=dict(img=114), scale=( 320, 320, ), type='LetterResize'), ], type='Compose'), dict( transforms=[ dict(scale=( 960, 960, ), type='YOLOv5KeepRatioResize'), dict( allow_scale_up=False, pad_val=dict(img=114), scale=( 960, 960, ), type='LetterResize'), ], type='Compose'), ], [ dict(prob=1.0, type='mmdet.RandomFlip'), dict(prob=0.0, type='mmdet.RandomFlip'), ], [ dict(type='mmdet.LoadAnnotations', with_bbox=True), ], [ dict( meta_keys=( 'img_id', 'img_path', 'ori_shape', 'img_shape', 'scale_factor', 'pad_param', 'flip', 'flip_direction', ), type='mmdet.PackDetInputs'), ], ], type='TestTimeAug'), ] val_ann_file = 'annotations/instances_val2017.json' val_batch_size_per_gpu = 1 val_cfg = dict(type='ValLoop') val_data_prefix = 'val2017/' val_dataloader = dict( batch_size=1, dataset=dict( ann_file='annotations/test.json', batch_shapes_cfg=None, data_prefix=dict(img='images/'), data_root='./data/coal_machine_20230817/', metainfo=dict(classes=('coal_machine', ), palette=[ ( 20, 220, 60, ), ]), pipeline=[ dict( file_client_args=dict(backend='disk'), type='LoadImageFromFile'), dict(scale=( 640, 640, ), type='YOLOv5KeepRatioResize'), dict( allow_scale_up=False, pad_val=dict(img=114), scale=( 640, 640, ), type='LetterResize'), dict(scope='mmdet', type='LoadAnnotations', with_bbox=True), dict( meta_keys=( 'img_id', 'img_path', 'ori_shape', 'img_shape', 'scale_factor', 'pad_param', ), type='mmdet.PackDetInputs'), ], test_mode=True, type='YOLOv5CocoDataset'), drop_last=False, num_workers=2, persistent_workers=True, pin_memory=True, sampler=dict(shuffle=False, type='DefaultSampler')) val_evaluator = dict( ann_file='./data/coal_machine_20230817/annotations/test.json', metric='bbox', proposal_nums=( 100, 1, 10, ), type='mmdet.CocoMetric') val_interval_stage2 = 1 val_num_workers = 2 vis_backends = [ dict(type='LocalVisBackend'), ] visualizer = dict( name='visualizer', type='mmdet.DetLocalVisualizer', vis_backends=[ dict(type='LocalVisBackend'), dict(type='TensorboardVisBackend'), ]) weight_decay = 0.0005 widen_factor = 0.5 work_dir = './work_dirs/yolov8_s_fast_1xb12-40e_coal_machine_20230817'

Additional information

是的

open-mmlab / mmyolo