search

open-mmlab / mmdetection

OpenMMLab Detection Toolbox and Benchmark
https://mmdetection.readthedocs.io
Apache License 2.0
29.21k stars 9.4k forks source link

CUDA error: device-side assert triggered (train a model include neg) #1815

Closed wendaomu closed 4 years ago

wendaomu commented 4 years ago

I use the code from dev/allow_empty_gt to train a model, and I have some negative sample, so when I make dataset in the coco format, I set the x,y,w,h, label 0. I have 6 classes to classify,so I set the num_classess 7.

I want to know whether the data I made produce the error?

python tools/train.py configs/my_cascade_rcnn_r50_fpn_1x.py --work_dir result/ 2019-12-15 09:30:21,276 - INFO - Distributed training: False 2019-12-15 09:30:21,276 - INFO - MMDetection Version: 1.0.rc0+unknown 2019-12-15 09:30:21,276 - INFO - Config: # model settings model = dict( type='CascadeRCNN', num_stages=3, pretrained='torchvision://resnet50', backbone=dict( type='ResNet', depth=50, num_stages=4, out_indices=(0, 1, 2, 3), frozen_stages=1, style='pytorch'), neck=dict( type='FPN', in_channels=[256, 512, 1024, 2048], out_channels=256, num_outs=5), rpn_head=dict( type='RPNHead', in_channels=256, feat_channels=256, anchor_scales=[8], anchor_ratios=[0.5, 1.0, 2.0], anchor_strides=[4, 8, 16, 32, 64], target_means=[.0, .0, .0, .0], target_stds=[1.0, 1.0, 1.0, 1.0], loss_cls=dict( type='CrossEntropyLoss', use_sigmoid=True, loss_weight=1.0), loss_bbox=dict(type='SmoothL1Loss', beta=1.0 / 9.0, loss_weight=1.0)), bbox_roi_extractor=dict( type='SingleRoIExtractor', roi_layer=dict(type='RoIAlign', out_size=7, sample_num=2), out_channels=256, featmap_strides=[4, 8, 16, 32]), bbox_head=[ dict( type='SharedFCBBoxHead', num_fcs=2, in_channels=256, fc_out_channels=1024, roi_feat_size=7, num_classes=7, target_means=[0., 0., 0., 0.], target_stds=[0.1, 0.1, 0.2, 0.2], reg_class_agnostic=True, loss_cls=dict( type='CrossEntropyLoss', use_sigmoid=False, loss_weight=1.0), loss_bbox=dict(type='SmoothL1Loss', beta=1.0, loss_weight=1.0)), dict( type='SharedFCBBoxHead', num_fcs=2, in_channels=256, fc_out_channels=1024, roi_feat_size=7, num_classes=7, target_means=[0., 0., 0., 0.], target_stds=[0.05, 0.05, 0.1, 0.1], reg_class_agnostic=True, loss_cls=dict( type='CrossEntropyLoss', use_sigmoid=False, loss_weight=1.0), loss_bbox=dict(type='SmoothL1Loss', beta=1.0, loss_weight=1.0)), dict( type='SharedFCBBoxHead', num_fcs=2, in_channels=256, fc_out_channels=1024, roi_feat_size=7, num_classes=7, target_means=[0., 0., 0., 0.], target_stds=[0.033, 0.033, 0.067, 0.067], reg_class_agnostic=True, loss_cls=dict( type='CrossEntropyLoss', use_sigmoid=False, loss_weight=1.0), loss_bbox=dict(type='SmoothL1Loss', beta=1.0, loss_weight=1.0)) ])

model training and testing settings

train_cfg = dict( rpn=dict( assigner=dict( type='MaxIoUAssigner', pos_iou_thr=0.7, neg_iou_thr=0.3, min_pos_iou=0.3, ignore_iof_thr=-1), sampler=dict( type='RandomSampler', num=256, pos_fraction=0.5, neg_pos_ub=-1, add_gt_as_proposals=False), allowed_border=0, pos_weight=-1, debug=False), rpn_proposal=dict( nms_across_levels=False, nms_pre=2000, nms_post=2000, max_num=2000, nms_thr=0.7, min_bbox_size=0), rcnn=[ dict( assigner=dict( type='MaxIoUAssigner', pos_iou_thr=0.5, neg_iou_thr=0.5, min_pos_iou=0.5, ignore_iof_thr=-1), sampler=dict( type='RandomSampler', num=512, pos_fraction=0.25, neg_pos_ub=-1, add_gt_as_proposals=True), pos_weight=-1, debug=False), dict( assigner=dict( type='MaxIoUAssigner', pos_iou_thr=0.6, neg_iou_thr=0.6, min_pos_iou=0.6, ignore_iof_thr=-1), sampler=dict( type='RandomSampler', num=512, pos_fraction=0.25, neg_pos_ub=-1, add_gt_as_proposals=True), pos_weight=-1, debug=False), dict( assigner=dict( type='MaxIoUAssigner', pos_iou_thr=0.7, neg_iou_thr=0.7, min_pos_iou=0.7, ignore_iof_thr=-1), sampler=dict( type='RandomSampler', num=512, pos_fraction=0.25, neg_pos_ub=-1, add_gt_as_proposals=True), pos_weight=-1, debug=False) ], stage_loss_weights=[1, 0.5, 0.25]) test_cfg = dict( rpn=dict( nms_across_levels=False, nms_pre=1000, nms_post=1000, max_num=1000, nms_thr=0.7, min_bbox_size=0), rcnn=dict( score_thr=0.05, nms=dict(type='nms', iou_thr=0.5), max_per_img=100), keep_all_stages=False)

dataset settings

dataset_type = 'CocoDataset' data_root = 'data/coco/' img_norm_cfg = dict( mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True) train_pipeline = [ dict(type='LoadImageFromFile'), dict(type='LoadAnnotations', with_bbox=True), dict(type='Resize', img_scale=(1333, 800), keep_ratio=True), dict(type='RandomFlip', flip_ratio=0.5), dict(type='Normalize', img_norm_cfg), dict(type='Pad', size_divisor=32), dict(type='DefaultFormatBundle'), dict(type='Collect', keys=['img', 'gt_bboxes', 'gt_labels']), ] test_pipeline = [ dict(type='LoadImageFromFile'), dict( type='MultiScaleFlipAug', img_scale=(1333, 800), flip=False, transforms=[ dict(type='Resize', keep_ratio=True), dict(type='RandomFlip'), dict(type='Normalize', img_norm_cfg), dict(type='Pad', size_divisor=32), dict(type='ImageToTensor', keys=['img']), dict(type='Collect', keys=['img']), ]) ] data = dict( imgs_per_gpu=4, workers_per_gpu=4, train=dict( type=dataset_type, ann_file=data_root + 'annotations/instances_train2017.json', img_prefix=data_root + 'images/train2017/', pipeline=train_pipeline), val=dict( type=dataset_type, ann_file=data_root + 'annotations/instances_val2017.json', img_prefix=data_root + 'images/val2017/', pipeline=test_pipeline), test=dict( type=dataset_type, ann_file=data_root + 'annotations/instances_val2017.json', img_prefix=data_root + 'images/val2017/', pipeline=test_pipeline))

optimizer

optimizer = dict(type='SGD', lr=0.0001, momentum=0.9, weight_decay=0.0001) optimizer_config = dict(grad_clip=dict(max_norm=35, norm_type=2))

learning policy

lr_config = dict( policy='step', warmup='linear', warmup_iters=500, warmup_ratio=1.0 / 10, step=[8, 11]) checkpoint_config = dict(interval=1)

yapf:disable

log_config = dict( interval=50, hooks=[ dict(type='TextLoggerHook'),

dict(type='TensorboardLoggerHook')

])

yapf:enable

runtime settings

total_epochs = 25 dist_params = dict(backend='nccl') log_level = 'INFO' work_dir = './work_dirs/cascade_rcnn_r50_fpn_1x' load_from = None resume_from = None workflow = [('train', 1)]

2019-12-15 09:30:21,998 - INFO - load model from: torchvision://resnet50 2019-12-15 09:30:30,103 - WARNING - The model and loaded state dict do not match exactly

unexpected key in source state_dict: fc.weight, fc.bias

loading annotations into memory... Done (t=1.19s) creating index... index created! 2019-12-15 09:30:41,608 - INFO - Start running, host: admin@fuxilabor_labor0_S4_Odps_S98_dsw_prepaid_cnbj_1181_201912080304, work_dir:/data/nas/workspace/jupyter/Project/lb_project/code/mmdetection-empty_gt/result 2019-12-15 09:30:41,608 - INFO - workflow: [('train', 1)], max: 25 epochs /pytorch/aten/src/THCUNN/ClassNLLCriterion.cu:57: void ClassNLLCriterion_updateOutput_no_reduce_kernel(int, THCDeviceTensor<Dtype, 2, int, DefaultPtrTraits>, THCDeviceTensor<long, 1, int, DefaultPtrTraits>, THCDeviceTensor<Dtype, 1, int, DefaultPtrTraits>, Dtype , int, int) [with Dtype = float]: block: [1,0,0], thread: [512,0,0] Assertion cur_target >= 0 && cur_target < n_classes failed. /pytorch/aten/src/THCUNN/ClassNLLCriterion.cu:57: void ClassNLLCriterion_updateOutput_no_reduce_kernel(int, THCDeviceTensor<Dtype, 2, int, DefaultPtrTraits>, THCDeviceTensor<long, 1, int, DefaultPtrTraits>, THCDeviceTensor<Dtype, 1, int, DefaultPtrTraits>, Dtype , int, int) [with Dtype = float]: block: [1,0,0], thread: [513,0,0] Assertion cur_target >= 0 && cur_target < n_classes failed. /pytorch/aten/src/THCUNN/ClassNLLCriterion.cu:57: void ClassNLLCriterion_updateOutput_no_reduce_kernel(int, THCDeviceTensor<Dtype, 2, int, DefaultPtrTraits>, THCDeviceTensor<long, 1, int, DefaultPtrTraits>, THCDeviceTensor<Dtype, 1, int, DefaultPtrTraits>, Dtype , int, int) [with Dtype = float]: block: [1,0,0], thread: [514,0,0] Assertion cur_target >= 0 && cur_target < n_classes failed. /pytorch/aten/src/THCUNN/ClassNLLCriterion.cu:57: void ClassNLLCriterion_updateOutput_no_reduce_kernel(int, THCDeviceTensor<Dtype, 2, int, DefaultPtrTraits>, THCDeviceTensor<long, 1, int, DefaultPtrTraits>, THCDeviceTensor<Dtype, 1, int, DefaultPtrTraits>, Dtype , int, int) [with Dtype = float]: block: [1,0,0], thread: [515,0,0] Assertion cur_target >= 0 && cur_target < n_classes failed. Traceback (most recent call last): File "tools/train.py", line 110, in main() File "tools/train.py", line 106, in main logger=logger) File "/home/admin/jupyter/Project/lb_project/code/mmdetection/mmdet/apis/train.py", line 60, in train_detector _non_dist_train(model, dataset, cfg, validate=validate) File "/home/admin/jupyter/Project/lb_project/code/mmdetection/mmdet/apis/train.py", line 232, in _non_dist_train runner.run(data_loaders, cfg.workflow, cfg.total_epochs) File "/data/nas/workspace/envs/python3.6/site-packages/mmcv/runner/runner.py", line 358, in run epoch_runner(data_loaders[i], kwargs) File "/data/nas/workspace/envs/python3.6/site-packages/mmcv/runner/runner.py", line 264, in train self.model, data_batch, train_mode=True, kwargs) File "/home/admin/jupyter/Project/lb_project/code/mmdetection/mmdet/apis/train.py", line 38, in batch_processor losses = model(data) File "/opt/conda/lib/python3.6/site-packages/torch/nn/modules/module.py", line 541, in call result = self.forward(*input, *kwargs) File "/opt/conda/lib/python3.6/site-packages/torch/nn/parallel/data_parallel.py", line 150, in forward return self.module(inputs[0], kwargs[0]) File "/opt/conda/lib/python3.6/site-packages/torch/nn/modules/module.py", line 541, in call result = self.forward(*input, *kwargs) File "/home/admin/jupyter/Project/lb_project/code/mmdetection/mmdet/core/fp16/decorators.py", line 49, in new_func return old_func(args, kwargs) File "/home/admin/jupyter/Project/lb_project/code/mmdetection/mmdet/models/detectors/base.py", line 117, in forward return self.forward_train(img, img_meta, kwargs) File "/home/admin/jupyter/Project/lb_project/code/mmdetection/mmdet/models/detectors/cascade_rcnn.py", line 247, in forward_train loss_bbox = bbox_head.loss(cls_score, bbox_pred, bbox_targets) File "/home/admin/jupyter/Project/lb_project/code/mmdetection/mmdet/core/fp16/decorators.py", line 127, in new_func return old_func(args, **kwargs) File "/home/admin/jupyter/Project/lb_project/code/mmdetection/mmdet/models/bbox_heads/bbox_head.py", line 120, in loss pos_bbox_pred = bbox_pred.view(bbox_pred.size(0), 4)[pos_inds] RuntimeError: copy_if failed to synchronize: device-side assert triggered terminate called after throwing an instance of 'c10::Error' what(): CUDA error: device-side assert triggered (insert_events at /pytorch/c10/cuda/CUDACachingAllocator.cpp:569) frame #0: c10::Error::Error(c10::SourceLocation, std::string const&) + 0x33 (0x7f3454350813 in /opt/conda/lib/python3.6/site-packages/torch/lib/libc10.so) frame #1: + 0x16126 (0x7f345458b126 in /opt/conda/lib/python3.6/site-packages/torch/lib/libc10_cuda.so) frame #2: + 0x16b11 (0x7f345458bb11 in /opt/conda/lib/python3.6/site-packages/torch/lib/libc10_cuda.so) frame #3: c10::TensorImpl::release_resources() + 0x4d (0x7f3454340f0d in /opt/conda/lib/python3.6/site-packages/torch/lib/libc10.so) frame #4: + 0x4af752 (0x7f3454dda752 in /opt/conda/lib/python3.6/site-packages/torch/lib/libtorch_python.so) frame #5: + 0x4af796 (0x7f3454dda796 in /opt/conda/lib/python3.6/site-packages/torch/lib/libtorch_python.so)

frame #25: __libc_start_main + 0xf5 (0x7f345fe22b15 in /lib64/libc.so.6) Aborted
yuyijie1995 commented 4 years ago

I got the same error: RuntimeError: tabulate: failed to synchronize: cudaErrorAssert: device-side assert triggered terminate called after throwing an instance of 'c10::Error'

After adding my own data aug method, model can be trained normally several epochs,then this error happened.

wendaomu commented 4 years ago

I set the label of neg image empty and then solve the problem.

hellock commented 4 years ago

Label 0 in annotation files does not mean negative, just leave it empty.

SunNYNO1 commented 4 years ago

hello,i meet the same questions, but i dont understand 'label 0' , how to leave it empty? i add some data to dataset, this is my xml: `<?xml version="1.0" ?>

2008_000032_2020.jpg 500 375 3 car 257 500 234 375 0 `
SunNYNO1 commented 4 years ago

能不能回答一下，刚开始以为是xml的格式不正确，修改后依然报同样的错误。并没有找到label为0的地方，我的类别如下，和pascal原本的类别是相同的： ['background', 'aeroplane', 'bicycle', 'bird', 'boat', 'bottle', 'bus', 'car', 'cat', 'chair', 'cow', 'diningtable', 'dog', 'horse', 'motorbike', 'person', 'pottedplant', 'sheep', 'sofa', 'train', 'tvmonitor'] 求解，抱拳，谢谢～

Beastmaster commented 4 years ago

Please check groundtruth label range and number of output channels. If label range is [0-2], you must output 3 channels to match the range. If label index 2 is to ignore, please pass "ignore_index=2" in NLLLoss, and you may output 2 channels.