TypeError: CocoDataset: __init__() got an unexpected keyword argument 'times'

Hi! Thanks for solid work. 👍

I have the following bug:

Description

I am receiving TypeError: CocoDataset: __init__() got an unexpected keyword argument 'times' while training from scratch CocoDataset like custom dataset. The same bug occurs when I also run SSD, DETR, YoloV3. But there is no issue with models, such as VFNet, Cascade, RetinaNet. So I cannot see clear reason for now.

Look forward to your help, Thank you! 🙏

Reproduction

I am running:

python centernet.py

where my centernet.py file is defined as below:

import torch
from mmcv import Config
from mmdet.apis import set_random_seed, train_detector
from mmdet.datasets import build_dataset
from mmdet.models import build_detector
from IPython.display import clear_output

cfg = Config.fromfile('./mmdetection/configs/centernet/centernet_resnet18_dcnv2_140e_coco.py')
DATASET_TYPE = 'CocoDataset'
PREFIX = '$path-to-my-images'
cfg.dataset_type = DATASET_TYPE
cfg.classes = ("CIN-1", "CIN-2", "CIN-3")

cfg.model.bbox_head.num_classes = 3

cfg.data.train.img_prefix = PREFIX
cfg.data.train.classes = cfg.classes
cfg.data.train.ann_file = '$path-to-train.json'
cfg.data.train.type = DATASET_TYPE

cfg.data.val.img_prefix = PREFIX
cfg.data.val.classes = cfg.classes
cfg.data.val.ann_file = '$path-to-val.json'
cfg.data.val.type = DATASET_TYPE

cfg.data.test.img_prefix = PREFIX
cfg.data.test.classes = cfg.classes
cfg.data.test.ann_file = '$path-to-test.json'
cfg.data.test.type = DATASET_TYPE

cfg.optimizer.lr = 0.01 / 8
cfg.optimizer_config.grad_clip = dict(max_norm=35, norm_type=2)
cfg.lr_config.policy = 'step'
cfg.lr_config.step = 7
#cfg.lr_config.warmup = None
#cfg.log_config.interval = 100

# Change the evaluation metric since we use customized dataset.
cfg.evaluation.metric = 'bbox'
# We can set the evaluation interval to reduce the evaluation
cfg.evaluation.interval = 4
# We can set the checkpoint saving interval to reduce the storage cost
cfg.checkpoint_config.interval = 4

# Set seed thus the results are more reproducible
cfg.seed = 0
set_random_seed(0, deterministic=False)
cfg.gpu_ids = range(1)

# we can use here mask_rcnn.
cfg.load_from = ''$path-to-pretrained-model.pth"
cfg.work_dir = ''$path-to-working-dir"

cfg.runner.max_epochs = 30
cfg.total_epochs = 30

clear_output()
model = build_detector(cfg.model)
datasets = [build_dataset(cfg.data.train)]

train_detector(model, datasets[0], cfg, distributed=False, validate=True)

Environment

1. Please run python mmdet/utils/collect_env.py to collect necessary environment information and paste it here.

sys.platform: linux
Python: 3.7.10 | packaged by conda-forge | (default, Feb 19 2021, 16:07:37) [GCC 9.3.0]
CUDA available: True
GPU 0,1,2,3,4,5: GeForce RTX 3090
CUDA_HOME: /usr/local/cuda-11.2
NVCC: Build cuda_11.2.r11.2/compiler.29558016_0
GCC: gcc (Ubuntu 9.3.0-17ubuntu1~20.04) 9.3.0
PyTorch: 1.9.0+cu111
PyTorch compiling details: PyTorch built with:
  - GCC 7.3
  - C++ Version: 201402
  - Intel(R) Math Kernel Library Version 2020.0.0 Product Build 20191122 for Intel(R) 64 architecture applications
  - Intel(R) MKL-DNN v2.1.2 (Git Hash 98be7e8afa711dc9b66c8ff3504129cb82013cdb)
  - OpenMP 201511 (a.k.a. OpenMP 4.5)
  - NNPACK is enabled
  - CPU capability usage: AVX2
  - CUDA Runtime 11.1
  - NVCC architecture flags: -gencode;arch=compute_37,code=sm_37;-gencode;arch=compute_50,code=sm_50;-gencode;arch=compute_60,code=sm_60;-gencode;arch=compute_70,code=sm_70;-gencode;arch=compute_75,code=sm_75;-gencode;arch=compute_80,code=sm_80;-gencode;arch=compute_86,code=sm_86
  - CuDNN 8.0.5
  - Magma 2.5.2
  - Build settings: BLAS_INFO=mkl, BUILD_TYPE=Release, CUDA_VERSION=11.1, CUDNN_VERSION=8.0.5, CXX_COMPILER=/opt/rh/devtoolset-7/root/usr/bin/c++, CXX_FLAGS= -Wno-deprecated -fvisibility-inlines-hidden -DUSE_PTHREADPOOL -fopenmp -DNDEBUG -DUSE_KINETO -DUSE_FBGEMM -DUSE_QNNPACK -DUSE_PYTORCH_QNNPACK -DUSE_XNNPACK -DSYMBOLICATE_MOBILE_DEBUG_HANDLE -O2 -fPIC -Wno-narrowing -Wall -Wextra -Werror=return-type -Wno-missing-field-initializers -Wno-type-limits -Wno-array-bounds -Wno-unknown-pragmas -Wno-sign-compare -Wno-unused-parameter -Wno-unused-variable -Wno-unused-function -Wno-unused-result -Wno-unused-local-typedefs -Wno-strict-overflow -Wno-strict-aliasing -Wno-error=deprecated-declarations -Wno-stringop-overflow -Wno-psabi -Wno-error=pedantic -Wno-error=redundant-decls -Wno-error=old-style-cast -fdiagnostics-color=always -faligned-new -Wno-unused-but-set-variable -Wno-maybe-uninitialized -fno-math-errno -fno-trapping-math -Werror=format -Wno-stringop-overflow, LAPACK_INFO=mkl, PERF_WITH_AVX=1, PERF_WITH_AVX2=1, PERF_WITH_AVX512=1, TORCH_VERSION=1.9.0, USE_CUDA=ON, USE_CUDNN=ON, USE_EXCEPTION_PTR=1, USE_GFLAGS=OFF, USE_GLOG=OFF, USE_MKL=ON, USE_MKLDNN=ON, USE_MPI=OFF, USE_NCCL=ON, USE_NNPACK=ON, USE_OPENMP=ON,

TorchVision: 0.10.0+cu111
OpenCV: 4.5.3
MMCV: 1.3.12
MMCV Compiler: GCC 7.3
MMCV CUDA Compiler: 11.1
MMDetection: 2.15.1+3545915

1. I installed mmdet with as follows:

• A.

conda create -n openmmlab python=3.7 -y
conda activate

• B.

pip install openmim
mim install mmdet