MaskTextSpotterV3报错
报错1
(MaskTextSpotterV3) xuehp@haomeiya004:~/git/MaskTextSpotterV3$ python tools/demo.py
Traceback (most recent call last):
File "tools/demo.py", line 6, in <module>
from maskrcnn_benchmark.modeling.detector import build_detection_model
File "/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/modeling/detector/__init__.py", line 2, in <module>
from .detectors import build_detection_model
File "/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/modeling/detector/detectors.py", line 3, in <module>
from .generalized_rcnn import GeneralizedRCNN
File "/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/modeling/detector/generalized_rcnn.py", line 12, in <module>
from ..backbone import build_backbone
File "/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/modeling/backbone/__init__.py", line 2, in <module>
from .backbone import build_backbone
File "/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/modeling/backbone/backbone.py", line 53, in <module>
from maskrcnn_benchmark.modeling.make_layers import conv_with_kaiming_uniform
File "/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/modeling/make_layers.py", line 10, in <module>
from maskrcnn_benchmark.layers import Conv2d
File "/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/layers/__init__.py", line 9, in <module>
from .nms import nms
File "/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/layers/nms.py", line 6, in <module>
from apex import amp
File "/home/xuehp/anaconda3/envs/MaskTextSpotterV3/lib/python3.7/site-packages/apex/__init__.py", line 13, in <module>
from pyramid.session import UnencryptedCookieSessionFactoryConfig
ImportError: cannot import name 'UnencryptedCookieSessionFactoryConfig' from 'pyramid.session' (unknown location)
原因：应该是apex安装失败导致的。
办法：重新安装apex；或者拷贝代码。
作者就是拷贝了maskrcnn_benchmark的代码，我删除了__init__.py文件，因为它引用了utils。
https://github.com/NVIDIA/apex
不支持mac
报错2
(MaskTextSpotterV3) xuehp@haomeiya008:~/git/MaskTextSpotterV3/tools$ python demo.py
Traceback (most recent call last):
File "demo.py", line 6, in <module>
[core]
from maskrcnn_benchmark.modeling.detector import build_detection_model
File "/home/xuehp/git/MaskTextSpotterV3/tools/maskrcnn_benchmark/modeling/detector/__init__.py", line 2, in <module>
from .detectors import build_detection_model
File "/home/xuehp/git/MaskTextSpotterV3/tools/maskrcnn_benchmark/modeling/detector/detectors.py", line 3, in <module>
from .generalized_rcnn import GeneralizedRCNN
File "/home/xuehp/git/MaskTextSpotterV3/tools/maskrcnn_benchmark/modeling/detector/generalized_rcnn.py", line 12, in <module>
from ..backbone import build_backbone
File "/home/xuehp/git/MaskTextSpotterV3/tools/maskrcnn_benchmark/modeling/backbone/__init__.py", line 2, in <module>
from .backbone import build_backbone
File "/home/xuehp/git/MaskTextSpotterV3/tools/maskrcnn_benchmark/modeling/backbone/backbone.py", line 53, in <module>
from maskrcnn_benchmark.modeling.make_layers import conv_with_kaiming_uniform
File "/home/xuehp/git/MaskTextSpotterV3/tools/maskrcnn_benchmark/modeling/make_layers.py", line 10, in <module>
from maskrcnn_benchmark.layers import Conv2d
File "/home/xuehp/git/MaskTextSpotterV3/tools/maskrcnn_benchmark/layers/__init__.py", line 9, in <module>
from .nms import nms
File "/home/xuehp/git/MaskTextSpotterV3/tools/maskrcnn_benchmark/layers/nms.py", line 3, in <module>
from maskrcnn_benchmark import _C
ImportError: /home/xuehp/git/MaskTextSpotterV3/tools/maskrcnn_benchmark/_C.cpython-36m-x86_64-linux-gnu.so: undefined symbol: _ZN2at7getTypeERKNS_6TensorE
这是因为torchvision没有安装好，可能是版本不对。
办法：重新安装torchvision
报错3
(MaskTextSpotterV3) xuehp@haomeiya008:~/git/MaskTextSpotterV3$ sh train.sh
*****************************************
Setting OMP_NUM_THREADS environment variable for each process to be 1 in default, to avoid your system being overloaded, please further tune the variable for optimal performance in your application as needed.
*****************************************
Traceback (most recent call last):
Traceback (most recent call last):
Traceback (most recent call last):
File "tools/train_net.py", line 19, in <module>
File "tools/train_net.py", line 19, in <module>
File "tools/train_net.py", line 19, in <module>
Traceback (most recent call last):
File "tools/train_net.py", line 19, in <module>
from maskrcnn_benchmark.modeling.detector import build_detection_modelfrom maskrcnn_benchmark.modeling.detector import build_detection_modelfrom maskrcnn_benchmark.modeling.detector import build_detection_model
File "/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/modeling/detector/__init__.py", line 2, in <module>
File "/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/modeling/detector/__init__.py", line 2, in <module>
File "/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/modeling/detector/__init__.py", line 2, in <module>
from maskrcnn_benchmark.modeling.detector import build_detection_model
File "/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/modeling/detector/__init__.py", line 2, in <module>
from .detectors import build_detection_model
File "/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/modeling/detector/detectors.py", line 3, in <module>
from .detectors import build_detection_model
from .detectors import build_detection_model File "/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/modeling/detector/detectors.py", line 3, in <module>
File "/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/modeling/detector/detectors.py", line 3, in <module>
from .detectors import build_detection_model
File "/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/modeling/detector/detectors.py", line 3, in <module>
from .generalized_rcnn import GeneralizedRCNN
File "/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/modeling/detector/generalized_rcnn.py", line 12, in <module>
from .generalized_rcnn import GeneralizedRCNN
File "/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/modeling/detector/generalized_rcnn.py", line 12, in <module>
from .generalized_rcnn import GeneralizedRCNN
from .generalized_rcnn import GeneralizedRCNN File "/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/modeling/detector/generalized_rcnn.py", line 12, in <module>
File "/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/modeling/detector/generalized_rcnn.py", line 12, in <module>
from ..backbone import build_backbone
File "/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/modeling/backbone/__init__.py", line 2, in <module>
from ..backbone import build_backbone
File "/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/modeling/backbone/__init__.py", line 2, in <module>
from ..backbone import build_backbone
File "/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/modeling/backbone/__init__.py", line 2, in <module>
from ..backbone import build_backbonefrom .backbone import build_backbone
File "/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/modeling/backbone/__init__.py", line 2, in <module>
File "/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/modeling/backbone/backbone.py", line 53, in <module>
from .backbone import build_backbone
File "/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/modeling/backbone/backbone.py", line 53, in <module>
from .backbone import build_backbone
File "/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/modeling/backbone/backbone.py", line 53, in <module>
from .backbone import build_backbone
File "/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/modeling/backbone/backbone.py", line 53, in <module>
from maskrcnn_benchmark.modeling.make_layers import conv_with_kaiming_uniform
File "/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/modeling/make_layers.py", line 10, in <module>
from maskrcnn_benchmark.modeling.make_layers import conv_with_kaiming_uniform
File "/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/modeling/make_layers.py", line 10, in <module>
from maskrcnn_benchmark.modeling.make_layers import conv_with_kaiming_uniform
File "/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/modeling/make_layers.py", line 10, in <module>
from maskrcnn_benchmark.modeling.make_layers import conv_with_kaiming_uniform
from maskrcnn_benchmark.layers import Conv2d File "/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/modeling/make_layers.py", line 10, in <module>
File "/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/layers/__init__.py", line 9, in <module>
from maskrcnn_benchmark.layers import Conv2d
File "/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/layers/__init__.py", line 9, in <module>
from maskrcnn_benchmark.layers import Conv2d
File "/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/layers/__init__.py", line 9, in <module>
from maskrcnn_benchmark.layers import Conv2d
File "/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/layers/__init__.py", line 9, in <module>
from .nms import nms
File "/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/layers/nms.py", line 3, in <module>
from .nms import nms
Traceback (most recent call last):
File "/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/layers/nms.py", line 3, in <module>
from .nms import nms File "tools/train_net.py", line 19, in <module>
File "/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/layers/nms.py", line 3, in <module>
from .nms import nms
from maskrcnn_benchmark import _C File "/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/layers/nms.py", line 3, in <module>
from maskrcnn_benchmark import _C
ImportError: cannot import name '_C' from 'maskrcnn_benchmark' (/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/__init__.py)ImportError
: from maskrcnn_benchmark import _Ccannot import name '_C' from 'maskrcnn_benchmark' (/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/__init__.py)
from maskrcnn_benchmark import _CImportError
: cannot import name '_C' from 'maskrcnn_benchmark' (/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/__init__.py)
ImportError: cannot import name '_C' from 'maskrcnn_benchmark' (/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/__init__.py)
from maskrcnn_benchmark.modeling.detector import build_detection_model
File "/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/modeling/detector/__init__.py", line 2, in <module>
from .detectors import build_detection_model
File "/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/modeling/detector/detectors.py", line 3, in <module>
from .generalized_rcnn import GeneralizedRCNN
File "/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/modeling/detector/generalized_rcnn.py", line 12, in <module>
from ..backbone import build_backbone
File "/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/modeling/backbone/__init__.py", line 2, in <module>
from .backbone import build_backbone
File "/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/modeling/backbone/backbone.py", line 53, in <module>
Traceback (most recent call last):
File "tools/train_net.py", line 19, in <module>
from maskrcnn_benchmark.modeling.make_layers import conv_with_kaiming_uniform
File "/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/modeling/make_layers.py", line 10, in <module>
from maskrcnn_benchmark.layers import Conv2d
File "/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/layers/__init__.py", line 9, in <module>
from .nms import nms
File "/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/layers/nms.py", line 3, in <module>
from maskrcnn_benchmark.modeling.detector import build_detection_model
File "/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/modeling/detector/__init__.py", line 2, in <module>
from maskrcnn_benchmark import _C
ImportError: cannot import name '_C' from 'maskrcnn_benchmark' (/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/__init__.py)
from .detectors import build_detection_model
File "/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/modeling/detector/detectors.py", line 3, in <module>
from .generalized_rcnn import GeneralizedRCNN
File "/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/modeling/detector/generalized_rcnn.py", line 12, in <module>
from ..backbone import build_backbone
File "/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/modeling/backbone/__init__.py", line 2, in <module>
from .backbone import build_backbone
File "/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/modeling/backbone/backbone.py", line 53, in <module>
from maskrcnn_benchmark.modeling.make_layers import conv_with_kaiming_uniform
File "/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/modeling/make_layers.py", line 10, in <module>
from maskrcnn_benchmark.layers import Conv2d
File "/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/layers/__init__.py", line 9, in <module>
from .nms import nms
File "/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/layers/nms.py", line 3, in <module>
from maskrcnn_benchmark import _C
ImportError: cannot import name '_C' from 'maskrcnn_benchmark' (/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/__init__.py)
Traceback (most recent call last):
File "tools/train_net.py", line 19, in <module>
from maskrcnn_benchmark.modeling.detector import build_detection_model
File "/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/modeling/detector/__init__.py", line 2, in <module>
from .detectors import build_detection_model
File "/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/modeling/detector/detectors.py", line 3, in <module>
from .generalized_rcnn import GeneralizedRCNN
File "/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/modeling/detector/generalized_rcnn.py", line 12, in <module>
from ..backbone import build_backbone
File "/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/modeling/backbone/__init__.py", line 2, in <module>
from .backbone import build_backbone
File "/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/modeling/backbone/backbone.py", line 53, in <module>
from maskrcnn_benchmark.modeling.make_layers import conv_with_kaiming_uniform
File "/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/modeling/make_layers.py", line 10, in <module>
from maskrcnn_benchmark.layers import Conv2d
File "/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/layers/__init__.py", line 9, in <module>
from .nms import nms
File "/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/layers/nms.py", line 3, in <module>
from maskrcnn_benchmark import _C
ImportError: cannot import name '_C' from 'maskrcnn_benchmark' (/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/__init__.py)
Traceback (most recent call last):
File "tools/train_net.py", line 19, in <module>
from maskrcnn_benchmark.modeling.detector import build_detection_model
File "/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/modeling/detector/__init__.py", line 2, in <module>
from .detectors import build_detection_model
File "/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/modeling/detector/detectors.py", line 3, in <module>
from .generalized_rcnn import GeneralizedRCNN
File "/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/modeling/detector/generalized_rcnn.py", line 12, in <module>
from ..backbone import build_backbone
File "/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/modeling/backbone/__init__.py", line 2, in <module>
from .backbone import build_backbone
File "/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/modeling/backbone/backbone.py", line 53, in <module>
from maskrcnn_benchmark.modeling.make_layers import conv_with_kaiming_uniform
File "/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/modeling/make_layers.py", line 10, in <module>
from maskrcnn_benchmark.layers import Conv2d
File "/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/layers/__init__.py", line 9, in <module>
from .nms import nms
File "/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/layers/nms.py", line 3, in <module>
from maskrcnn_benchmark import _C
ImportError: cannot import name '_C' from 'maskrcnn_benchmark' (/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/__init__.py)
Traceback (most recent call last):
File "/home/xuehp/anaconda3/envs/MaskTextSpotterV3/lib/python3.8/runpy.py", line 194, in _run_module_as_main
return _run_code(code, main_globals, None,
File "/home/xuehp/anaconda3/envs/MaskTextSpotterV3/lib/python3.8/runpy.py", line 87, in _run_code
exec(code, run_globals)
File "/home/xuehp/anaconda3/envs/MaskTextSpotterV3/lib/python3.8/site-packages/torch/distributed/launch.py", line 263, in <module>
main()
File "/home/xuehp/anaconda3/envs/MaskTextSpotterV3/lib/python3.8/site-packages/torch/distributed/launch.py", line 258, in main
raise subprocess.CalledProcessError(returncode=process.returncode,
subprocess.CalledProcessError: Command '['/home/xuehp/anaconda3/envs/MaskTextSpotterV3/bin/python', '-u', 'tools/train_net.py', '--local_rank=7', '--config-file', 'configs/pretrain/seg_rec_poly_fuse_feature.yaml\r']' returned non-zero exit status 1.
(MaskTextSpotterV3) xuehp@haomeiya008:~/git/MaskTextSpotterV3$
办法：直接在命令行敲入运行命令及参数
(MaskTextSpotterV3) xuehp@haomeiya008:~/git/MaskTextSpotterV3$ python tools/train_net.py --config-file configs/pretrain/seg_rec_poly_fuse_feature.yaml
报错4
(MaskTextSpotterV3) xuehp@haomeiya008:~/git/MaskTextSpotterV3$ python tools/train_net.py --config-file configs/pretrain/seg_rec_poly_fuse_feature.yaml
2021-03-12 13:01:12,664 maskrcnn_benchmark INFO: Using 1 GPUs
2021-03-12 13:01:12,664 maskrcnn_benchmark INFO: Namespace(config_file='configs/pretrain/seg_rec_poly_fuse_feature.yaml', distributed=False, local_rank=0, opts=[], skip_test=False)
2021-03-12 13:01:12,664 maskrcnn_benchmark INFO: Collecting env info (might take some time)
2021-03-12 13:01:14,345 maskrcnn_benchmark INFO:
PyTorch version: 1.4.0
Is debug build: No
CUDA used to build PyTorch: 10.0
OS: Ubuntu 18.04.5 LTS
GCC version: (Ubuntu 7.5.0-3ubuntu1~18.04) 7.5.0
CMake version: version 3.10.2
Python version: 3.8
Is CUDA available: Yes
CUDA runtime version: Could not collect
GPU models and configuration:
GPU 0: Tesla T4
GPU 1: Tesla T4
Nvidia driver version: 440.64.00
cuDNN version: /usr/local/cuda-10.1/targets/x86_64-linux/lib/libcudnn.so.7.6.5
Versions of relevant libraries:
[pip3] numpy==1.19.2
[pip3] torch==1.4.0
[pip3] torchvision==0.5.0
[conda] blas 1.0 mkl
[conda] mkl 2020.2 256
[conda] mkl-service 2.3.0 py38he904b0f_0
[conda] mkl_fft 1.3.0 py38h54f3939_0
[conda] mkl_random 1.1.1 py38h0573a6f_0
[conda] pytorch 1.4.0 py3.8_cuda10.0.130_cudnn7.6.3_0 pytorch
[conda] torchvision 0.5.0 py38_cu100 pytorch
Pillow (8.1.2)
2021-03-12 13:01:14,345 maskrcnn_benchmark INFO: Loaded configuration file configs/pretrain/seg_rec_poly_fuse_feature.yaml
2021-03-12 13:01:14,345 maskrcnn_benchmark INFO:
MODEL:
META_ARCHITECTURE: "GeneralizedRCNN"
WEIGHT: "catalog://ImageNetPretrained/MSRA/R-50"
BACKBONE:
CONV_BODY: "R-50-FPN"
OUT_CHANNELS: 256
RESNETS:
BACKBONE_OUT_CHANNELS: 256
RPN:
USE_FPN: True
ANCHOR_STRIDE: (4, 8, 16, 32, 64)
PRE_NMS_TOP_N_TRAIN: 2000
PRE_NMS_TOP_N_TEST: 1000
POST_NMS_TOP_N_TEST: 1000
FPN_POST_NMS_TOP_N_TEST: 1000
SEG:
USE_FPN: True
USE_FUSE_FEATURE: True
TOP_N_TRAIN: 1000
TOP_N_TEST: 1000
BINARY_THRESH: 0.1
BOX_THRESH: 0.1
MIN_SIZE: 5
SHRINK_RATIO: 0.4
EXPAND_RATIO: 3.0
ROI_HEADS:
USE_FPN: True
BATCH_SIZE_PER_IMAGE: 512
ROI_BOX_HEAD:
POOLER_RESOLUTION: 7
POOLER_SCALES: (0.25,)
POOLER_SAMPLING_RATIO: 2
FEATURE_EXTRACTOR: "FPN2MLPFeatureExtractor"
PREDICTOR: "FPNPredictor"
NUM_CLASSES: 2
USE_MASKED_FEATURE: True
ROI_MASK_HEAD:
POOLER_SCALES: (0.25,)
FEATURE_EXTRACTOR: "MaskRCNNFPNFeatureExtractor"
PREDICTOR: "SeqCharMaskRCNNC4Predictor"
POOLER_RESOLUTION: 14
POOLER_RESOLUTION_H: 32
POOLER_RESOLUTION_W: 32
POOLER_SAMPLING_RATIO: 2
RESOLUTION: 28
RESOLUTION_H: 64
RESOLUTION_W: 64
SHARE_BOX_FEATURE_EXTRACTOR: False
CHAR_NUM_CLASSES: 37
USE_WEIGHTED_CHAR_MASK: True
MASK_BATCH_SIZE_PER_IM: 64
USE_MASKED_FEATURE: True
MASK_ON: True
CHAR_MASK_ON: True
SEG_ON: True
SEQUENCE:
SEQ_ON: True
NUM_CHAR: 38
BOS_TOKEN: 0
MAX_LENGTH: 32
TEACHER_FORCE_RATIO: 1.0
DATASETS:
TRAIN: ("synthtext_train",)
# TRAIN: ("synthtext_train","icdar_2013_train","icdar_2015_train","scut-eng-char_train","total_text_train")
# RATIOS: [0.25,0.25,0.25,0.125,0.125]
TEST: ("icdar_2015_test",)
# TEST: ("total_text_test",)
AUG: True
IGNORE_DIFFICULT: True
MAX_ROTATE_THETA: 90
DATALOADER:
SIZE_DIVISIBILITY: 32
NUM_WORKERS: 4
ASPECT_RATIO_GROUPING: False
SOLVER:
BASE_LR: 0.02 #0.02
WARMUP_FACTOR: 0.1
WEIGHT_DECAY: 0.0001
STEPS: (100000, 200000)
MAX_ITER: 300000
IMS_PER_BATCH: 8
RESUME: True
DISPLAY_FREQ: 20
OUTPUT_DIR: "./output/pretrain"
TEST:
VIS: False
CHAR_THRESH: 192
IMS_PER_BATCH: 1
INPUT:
MIN_SIZE_TRAIN: (600, 800)
# MIN_SIZE_TRAIN: (800, 1000, 1200, 1400)
MAX_SIZE_TRAIN: 2333
MIN_SIZE_TEST: 1440
MAX_SIZE_TEST: 4000
2021-03-12 13:01:14,346 maskrcnn_benchmark INFO: Running with config:
AMP_VERBOSE: False
DATALOADER:
ASPECT_RATIO_GROUPING: False
NUM_WORKERS: 4
SIZE_DIVISIBILITY: 32
DATASETS:
AUG: True
CROP_SIZE: (512, 512)
FIX_CROP: False
FIX_ROTATE: False
IGNORE_DIFFICULT: True
MAX_ROTATE_THETA: 90
RANDOM_CROP_PROB: 0.0
RATIOS: []
TEST: ('icdar_2015_test',)
TRAIN: ('synthtext_train',)
DTYPE: float32
INPUT:
MAX_SIZE_TEST: 4000
MAX_SIZE_TRAIN: 2333
MIN_SIZE_TEST: 1440
MIN_SIZE_TRAIN: (600, 800)
PIXEL_MEAN: [102.9801, 115.9465, 122.7717]
PIXEL_STD: [1.0, 1.0, 1.0]
STRICT_RESIZE: False
TO_BGR255: True
MODEL:
BACKBONE:
CONV_BODY: R-50-FPN
FREEZE_CONV_BODY_AT: 2
OUT_CHANNELS: 256
CHAR_MASK_ON: True
DEVICE: cuda
FPN:
USE_GN: False
USE_RELU: False
MASK_ON: True
META_ARCHITECTURE: GeneralizedRCNN
RESNET34: False
RESNETS:
BACKBONE_OUT_CHANNELS: 256
DEFORMABLE_GROUPS: 1
LAYERS: (3, 4, 6, 3)
NUM_GROUPS: 1
RES2_OUT_CHANNELS: 256
RES5_DILATION: 1
STAGE_WITH_DCN: (False, False, False, False)
STEM_FUNC: StemWithFixedBatchNorm
STEM_OUT_CHANNELS: 64
STRIDE_IN_1X1: True
TRANS_FUNC: BottleneckWithFixedBatchNorm
WIDTH_PER_GROUP: 64
WITH_MODULATED_DCN: False
ROI_BOX_HEAD:
FEATURE_EXTRACTOR: FPN2MLPFeatureExtractor
INFERENCE_USE_BOX: True
MIX_OPTION:
MLP_HEAD_DIM: 1024
NUM_CLASSES: 2
POOLER_RESOLUTION: 7
POOLER_SAMPLING_RATIO: 2
POOLER_SCALES: (0.25,)
PREDICTOR: FPNPredictor
SOFT_MASKED_FEATURE_RATIO: 0.0
USE_MASKED_FEATURE: True
USE_REGRESSION: True
ROI_HEADS:
BATCH_SIZE_PER_IMAGE: 512
BBOX_REG_WEIGHTS: (10.0, 10.0, 5.0, 5.0)
BG_IOU_THRESHOLD: 0.5
DETECTIONS_PER_IMG: 100
FG_IOU_THRESHOLD: 0.5
NMS: 0.5
POSITIVE_FRACTION: 0.25
SCORE_THRESH: 0.0
USE_FPN: True
ROI_MASK_HEAD:
CHAR_NUM_CLASSES: 37
CONV_LAYERS: (256, 256, 256, 256)
FEATURE_EXTRACTOR: MaskRCNNFPNFeatureExtractor
MASK_BATCH_SIZE_PER_IM: 64
MIX_OPTION:
MLP_HEAD_DIM: 1024
POOLER_RESOLUTION: 14
POOLER_RESOLUTION_H: 32
POOLER_RESOLUTION_W: 32
POOLER_SAMPLING_RATIO: 2
POOLER_SCALES: (0.25,)
PREDICTOR: SeqCharMaskRCNNC4Predictor
RESOLUTION: 28
RESOLUTION_H: 64
RESOLUTION_W: 64
SHARE_BOX_FEATURE_EXTRACTOR: False
SOFT_MASKED_FEATURE_RATIO: 0.0
USE_MASKED_FEATURE: True
USE_WEIGHTED_CHAR_MASK: True
RPN:
ANCHOR_SIZES: (32, 64, 128, 256, 512)
ANCHOR_STRIDE: (4, 8, 16, 32, 64)
ASPECT_RATIOS: (0.5, 1.0, 2.0)
BATCH_SIZE_PER_IMAGE: 256
BG_IOU_THRESHOLD: 0.3
FG_IOU_THRESHOLD: 0.7
FPN_POST_NMS_TOP_N_TEST: 1000
FPN_POST_NMS_TOP_N_TRAIN: 2000
MIN_SIZE: 0
NMS_THRESH: 0.7
POSITIVE_FRACTION: 0.5
POST_NMS_TOP_N_TEST: 1000
POST_NMS_TOP_N_TRAIN: 2000
PRE_NMS_TOP_N_TEST: 1000
PRE_NMS_TOP_N_TRAIN: 2000
STRADDLE_THRESH: 0
USE_FPN: True
RPN_ONLY: False
SEG:
AUG_PROPOSALS: False
BATCH_SIZE_PER_IMAGE: 256
BINARY_THRESH: 0.1
BOX_EXPAND_RATIO: 1.5
BOX_THRESH: 0.1
EXPAND_RATIO: 3.0
IGNORE_DIFFICULT: True
MIN_SIZE: 5
MULTIPLE_THRESH: (0.2, 0.3, 0.5, 0.7)
POSITIVE_FRACTION: 0.5
SHRINK_RATIO: 0.4
TOP_N_TEST: 1000
TOP_N_TRAIN: 1000
USE_FPN: True
USE_FUSE_FEATURE: True
USE_MULTIPLE_THRESH: False
USE_PPM: False
USE_SEG_POLY: False
SEG_ON: True
TRAIN_DETECTION_ONLY: False
WEIGHT: catalog://ImageNetPretrained/MSRA/R-50
OUTPUT_DIR: ./output/pretrain
PATHS_CATALOG: /home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/config/paths_catalog.py
SEQUENCE:
BOS_TOKEN: 0
MAX_LENGTH: 32
MEAN_SCORE: False
NUM_CHAR: 38
RESIZE_HEIGHT: 16
RESIZE_WIDTH: 64
SEQ_ON: True
TEACHER_FORCE_RATIO: 1.0
TWO_CONV: False
SOLVER:
BASE_LR: 0.02
BIAS_LR_FACTOR: 2
CHECKPOINT_PERIOD: 5000
DISPLAY_FREQ: 20
GAMMA: 0.1
IMS_PER_BATCH: 8
MAX_ITER: 300000
MOMENTUM: 0.9
POW_SCHEDULE: False
RESUME: True
STEPS: (100000, 200000)
USE_ADAM: False
WARMUP_FACTOR: 0.1
WARMUP_ITERS: 500
WARMUP_METHOD: linear
WEIGHT_DECAY: 0.0001
WEIGHT_DECAY_BIAS: 0
TEST:
CHAR_THRESH: 192
EXPECTED_RESULTS: []
EXPECTED_RESULTS_SIGMA_TOL: 4
IMS_PER_BATCH: 1
VIS: False
Selected optimization level O0: Pure FP32 training.
Defaults for this optimization level are:
enabled : True
opt_level : O0
cast_model_type : torch.float32
patch_torch_functions : False
keep_batchnorm_fp32 : None
master_weights : False
loss_scale : 1.0
Processing user overrides (additional kwargs that are not None)...
After processing overrides, optimization options are:
enabled : True
opt_level : O0
cast_model_type : torch.float32
patch_torch_functions : False
keep_batchnorm_fp32 : None
master_weights : False
loss_scale : 1.0
Warning: multi_tensor_applier fused unscale kernel is unavailable, possibly because apex was installed without --cuda_ext --cpp_ext. Using Python fallback. Original ImportError was: ModuleNotFoundError("No module named 'amp_C'")
2021-03-12 13:01:17,745 maskrcnn_benchmark.utils.checkpoint INFO: Loading checkpoint from catalog://ImageNetPretrained/MSRA/R-50
2021-03-12 13:01:17,745 maskrcnn_benchmark.utils.checkpoint INFO: catalog://ImageNetPretrained/MSRA/R-50 points to https://dl.fbaipublicfiles.com/detectron/ImageNetPretrained/MSRA/R-50.pkl
2021-03-12 13:01:17,745 maskrcnn_benchmark.utils.checkpoint INFO: url https://dl.fbaipublicfiles.com/detectron/ImageNetPretrained/MSRA/R-50.pkl cached in /home/xuehp/.torch/models/R-50.pkl
/home/xuehp/.torch/models/R-50.pkl
Traceback (most recent call last):
File "tools/train_net.py", line 153, in <module>
main()
File "tools/train_net.py", line 149, in main
model = train(cfg, args.local_rank, args.distributed)
File "tools/train_net.py", line 63, in train
extra_checkpoint_data = checkpointer.load(cfg.MODEL.WEIGHT, resume=cfg.SOLVER.RESUME)
File "/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/utils/checkpoint.py", line 61, in load
checkpoint = self._load_file(f)
File "/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/utils/checkpoint.py", line 134, in _load_file
return load_c2_format(self.cfg, f)
File "/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/utils/c2_model_loading.py", line 164, in load_c2_format
state_dict = _load_c2_pickled_weights(f)
File "/home/xuehp/git/MaskTextSpotterV3/maskrcnn_benchmark/utils/c2_model_loading.py", line 123, in _load_c2_pickled_weights
data = pickle.load(f, encoding="latin1")
UnicodeDecodeError: 'latin1' codec can't decode byte 0x95 in position 6: illegal multibyte sequence
(MaskTextSpotterV3) xuehp@haomeiya008:~/git/MaskTextSpotterV3$
办法：这个文件下载错了，重新下载，https://dl.fbaipublicfiles.com/detectron/ImageNetPretrained/MSRA/R-50.pkl
放置到：/home/xuehp/.torch/models/R-50.pkl
运行起来了
（1）将dataset目录下的文件进行解压
(MaskTextSpotterV3) xuehp@haomeiya008:~/git/MaskTextSpotterV3$ ls datasets/
icdar2013 icdar2015 scut-eng-char SynthText_GT_E2E total_text_labels
icdar2013.zip icdar2015.zip scut-eng-char.zip SynthText_GT_E2E.tar.gz total_text_labels.zip
（2）修改了配置文件
DATASETS:
TRAIN: ("icdar_2015_train",) #这块儿和测试文件保持一样
TEST: ("icdar_2015_test",)
（3）运行
(MaskTextSpotterV3) xuehp@haomeiya008:~/git/MaskTextSpotterV3$ python tools/train_net.py --config-file configs/pretrain/seg_rec_poly_fuse_feature.yaml
失败了
一张显卡不够用
