Support PointRend, Fast_SCNN, HRNet, Deeplabv3_plus(xception, resnet, mobilenet), ContextNet, FPENet, DABNet, EdaNet, ENet, Espnetv2, RefineNet, UNet, DANet, HRNet, DFANet, HardNet, LedNet, OCNet, EncNet, DuNet, CGNet, CCNet, BiSeNet, PSPNet, ICNet, FCN, deeplab)
License: Apache License 2.0
conda create --name pointrend python=3.6
conda install pytorch==1.1.0 torchvision==0.3.0 cudatoolkit=9.0 -c pytorch
CUDA_VISIBLE_DEVICES=2 python -u tools/train.py --config-file configs/pascal_voc_deeplabv3_plus.yaml
[THOP Defined] Conv2d
[THOP Defined] BatchNorm2d
[THOP Defined] ReLU
[THOP Defined] Conv2d
[THOP Defined] BatchNorm2d
[THOP Defined] ReLU
[THOP Defined] Conv2d
[THOP Defined] BatchNorm2d
[THOP Defined] ReLU
[THOP Defined] Conv2d
[THOP Defined] BatchNorm2d
[THOP Defined] ReLU
[THOP Defined] Conv2d
<class 'torch.Tensor'>
2020-04-10 15:14:51,053 Segmentron WARNING: get flops and params error: new() received an invalid combination of arguments - got (list, dtype=torch.dtype), but expected one of:
2020-04-10 15:14:51,054 Segmentron INFO: Not use SyncBatchNorm!
2020-04-10 15:14:51,054 Segmentron INFO: Set bn custom eps for bn in encoder: 0.001
2020-04-10 15:14:51,060 Segmentron INFO: Start training, Total Epochs: 50 = Total Iterations 9150
<class 'torch.Tensor'>
Traceback (most recent call last):
File "tools/train.py", line 221, in
trainer.train()
File "tools/train.py", line 133, in train
outputs = self.model(images)
File "/home/yangwx/anaconda2/envs/pointrend/lib/python3.6/site-packages/torch/nn/modules/module.py", line 493, in call
result = self.forward(*input, **kwargs)
File "/mnt/data2/yangwx/GIT/SegmenTron/segmentron/models/deeplabv3_plus.py", line 36, in forward
c1, _, c3, c4 = self.encoder(x)
File "/home/yangwx/anaconda2/envs/pointrend/lib/python3.6/site-packages/torch/nn/modules/module.py", line 493, in call
result = self.forward(*input, **kwargs)
File "/mnt/data2/yangwx/GIT/SegmenTron/segmentron/models/backbones/xception.py", line 131, in forward
x = self.conv1(x)
File "/home/yangwx/anaconda2/envs/pointrend/lib/python3.6/site-packages/torch/nn/modules/module.py", line 495, in call
hook_result = hook(self, input, result)
File "/home/yangwx/anaconda2/envs/pointrend/lib/python3.6/site-packages/thop/vision/basic_hooks.py", line 27, in count_convNd
m.total_ops += torch.DoubleTensor([int(total_ops)], dtype=torch.float64)
TypeError: new() received an invalid combination of arguments - got (list, dtype=torch.dtype), but expected one of:
I train segmentation mode (deeplabV3+MobileNetV2) on my own dataset, demo.py can run , but the result was wrong. all of the catergory is 0(backgourd).
https://github.com/LikeLy-Journey/SegmenTron/blob/master/tools/demo.py
backbone: Inception is ok
backbone: MobileNet is not ok .
but if I add below source code in demo.py , it will be work.
if hasattr(self.model, 'encoder') and hasattr(self.model.encoder, 'named_modules') and \
cfg.MODEL.BN_EPS_FOR_ENCODER:
logging.info('set bn custom eps for bn in encoder: {}'.format(cfg.MODEL.BN_EPS_FOR_ENCODER))
self.set_batch_norm_attr(self.model.encoder.named_modules(), 'eps', cfg.MODEL.BN_EPS_FOR_ENCODER)
It will change EPS , not use the default.
And I found in https://github.com/LikeLy-Journey/SegmenTron/blob/master/segmentron/solver/optimizer.py
def _get_paramters(model):
params_list = list()
if hasattr(model, 'encoder') and model.encoder is not None and hasattr(model, 'decoder'):
params_list.append({'params': model.encoder.parameters(), 'lr': cfg.SOLVER.LR})
if cfg.MODEL.BN_EPS_FOR_ENCODER:
logging.info('Set bn custom eps for bn in encoder: {}'.format(cfg.MODEL.BN_EPS_FOR_ENCODER))
_set_batch_norm_attr(model.encoder.named_modules(), 'eps', cfg.MODEL.BN_EPS_FOR_ENCODER)
I am confused about this change of EPS.
would you please explain the reason of change BN EPS ?
RuntimeError: Attempting to deserialize object on a CUDA device but torch.cuda.is_available() is False.
If you are running on a CPU-only machine, please use torch.load with map_location='cpu' to map your storages to the CPU.
fix: model_zoo.py
add: map_location=cpu or cuda
torch.load(cfg.TRAIN.PRETRAINED_MODEL_PATH, map_location=torch.device(cfg.DEVICE))
Thanks for your nice work!
How to generate the prediction files on cityscapes test set and simply zip the prediction folder and submit to the Cityscapes submission page
hope for your help, thanks
You use the resolution of 769 * 769 when training cityscapes but use 1025 * 2049 when validation。
If there be a gap between the two results
Without any change, I train with 8 gpus but only get 59.0 mIoU on cityscapes. But using your trained fast_scnn_segmentron.pth can reproduce 68.9
Did you change something later about the config yaml?
Thanks for your nice work!
Thank you for your project. We deployed this project on the newly installed 8GPU server and intend to run all the demos. We did not change the configs and tested three demos: DeepLabv3_plus_resnet101_cityscape (≈78.28), DFANet_xceptionA (≈59.10), Fast_SCNN (≈60.08), the latter two are far from the original results, so the experiment was stopped.
For some well-known reasons, the results of the AI paper are not easy to reproduce, but we noticed that you mentioned in a previous reply that the configuration of a demo needs to be modified mode = 'testval', whether each demo should modify some code in order to reach the mIoU listed in the original text or the project? Or every configs need not change,the low mIoU just because the original results are difficult to reproduce?
I did not see the enet you train.
Have you train the enet model?
Thank you for your great job! Is there any pretrained model for DFANet?
I want to train my own dataset based on cityscapes, what kind of structures the folds of my dataset need satisfy? Thank you. My wechat is 18810220665
Hello Sir,
I tested and checked your code.
But, I could not find reference paper of your networks.
(Some networks, I could find reference paper.
for example, FastSCNN :
Chen, Liang-Chieh, et al. "Encoder-Decoder with Atrous Separable Convolution for Semantic Image Segmentation )
If you don't mind, please tell reference site or paper for your networks..
(Not found networks : HRNet, DFANet, ICNet, EncNet, HardNet, DABNet, FPENet, ContextNet, ENet, EDANet, PointRend)
Thanks..
Hi,
In espnetv2 and enet models I am troubling to find the classifier head for adjusting it to my DataSet.
Thank you very much for the implementation of all the models.
How do you test the speed of different segmentation models, sometimes, I am a little confused about the speed of some papers description？
Thanks for your nice work~
Can you provide an imagenet pre-trained script? Such as training mobilenetv2 backbone for deeplabv3 plus
python ./tools/eval.py --config-file configs/pascal_voc_deeplabv3_plus.yaml \
TEST.TEST_MODEL_PATH weights/deeplabv3_plus_xception_pascal_aug_segmentron.pth \
TEST.BATCH_SIZE 1
2020-03-18 10:41:21---Congratulations, the task is accomplished
2020-03-18 10:41:20---The training is finished, and output file is producing .........
2020-03-18 10:41:20---The training is completed and the files are transferred.
2020-03-18 10:41:09---| 4 | boat | 0.694096 |
2020-03-18 10:41:09---+------------+--------------+----------+
2020-03-18 10:41:09---| 5 | bottle | 0.889787 |
2020-03-18 10:41:09---+------------+--------------+----------+
2020-03-18 10:41:09---| 6 | bus | 0.946816 |
2020-03-18 10:41:09---+------------+--------------+----------+
2020-03-18 10:41:09---| 7 | car | 0.910226 |
2020-03-18 10:41:09---+------------+--------------+----------+
2020-03-18 10:41:09---| 8 | cat | 0.934975 |
2020-03-18 10:41:09---+------------+--------------+----------+
2020-03-18 10:41:09---| 9 | chair | 0.578426 |
2020-03-18 10:41:09---+------------+--------------+----------+
2020-03-18 10:41:09---| 10 | cow | 0.941479 |
2020-03-18 10:41:09---+------------+--------------+----------+
2020-03-18 10:41:09---| 11 | diningtable | 0.657297 |
2020-03-18 10:41:09---+------------+--------------+----------+
2020-03-18 10:41:09---| 12 | dog | 0.920236 |
2020-03-18 10:41:09---+------------+--------------+----------+
2020-03-18 10:41:09---| 13 | horse | 0.918536 |
2020-03-18 10:41:09---+------------+--------------+----------+
2020-03-18 10:41:09---| 14 | motorcycle | 0.885518 |
2020-03-18 10:41:09---+------------+--------------+----------+
2020-03-18 10:41:09---| 15 | person | 0.882771 |
2020-03-18 10:41:09---+------------+--------------+----------+
2020-03-18 10:41:09---| 16 | potted-plant | 0.739484 |
2020-03-18 10:41:09---+------------+--------------+----------+
2020-03-18 10:41:09---| 17 | sheep | 0.919957 |
2020-03-18 10:41:09---+------------+--------------+----------+
2020-03-18 10:41:09---| 18 | sofa | 0.723864 |
2020-03-18 10:41:09---+------------+--------------+----------+
2020-03-18 10:41:09---| 19 | train | 0.886305 |
2020-03-18 10:41:09---+------------+--------------+----------+
2020-03-18 10:41:09---| 20 | tv | 0.797698 |
2020-03-18 10:41:09---+------------+--------------+----------+
2020-03-18 10:41:09---2020-03-18 10:41:09,562 Segmentron INFO: Category iou:
2020-03-18 10:41:09---+------------+--------------+----------+
2020-03-18 10:41:09---| class id | class name | iou |
2020-03-18 10:41:09---+============+==============+==========+
2020-03-18 10:41:09---| 0 | background | 0.952956 |
2020-03-18 10:41:09---+------------+--------------+----------+
2020-03-18 10:41:09---| 1 | airplane | 0.882033 |
2020-03-18 10:41:09---+------------+--------------+----------+
2020-03-18 10:41:09---| 2 | bicycle | 0.412242 |
2020-03-18 10:41:09---+------------+--------------+----------+
2020-03-18 10:41:09---| 3 | bird | 0.892831 |
2020-03-18 10:41:09---+------------+--------------+----------+
2020-03-18 10:41:09---2020-03-18 10:41:09,554 Segmentron INFO: Eval use time: 453.290 second
2020-03-18 10:41:09---2020-03-18 10:41:09,555 Segmentron INFO: End validation pixAcc: 95.921, mIoU: 82.703
Thanks for your great job!
I wonder if there any visualization work for PointRend.
I wanna present the improvement due to the introduction of PointRend, just like the paper mentioned.
Traceback (most recent call last):
File "./tools/train.py", line 224, in <module>
trainer.train()
File "./tools/train.py", line 166, in train
self.validation(epoch)
File "./tools/train.py", line 198, in validation
self.metric.update(output, target)
File "/SegmenTron/segmentron/utils/score.py", line 58, in update
evaluate_worker(self, preds, labels)
File "/SegmenTron/segmentron/utils/score.py", line 41, in evaluate_worker
correct, labeled = batch_pix_accuracy(pred, label)
File "/SegmenTron/segmentron/utils/score.py", line 93, in batch_pix_accuracy
pixel_correct = torch.sum((predict == target) * (target > 0))#.item()
RuntimeError: The size of tensor a (2040) must match the size of tensor b (2048) at non-singleton dimension 2
when I run ./tools/dist_train.sh configs/cityscapes_icnet_resnet.yaml 4, the above error was generated during validation
Hello Sir,
I was tested your code using my data (image resolution : 256).
But I could not train many cases.
(unfortunately, I could not train all backbone of DFANet, PSPNet, ICNet)
If you don't mind, please tell me trainable networks and matched backbone.
Thanks.
In {$PYTHON -m torch.distributed.launch --nproc_per_node=$GPUS
$(dirname "$0")/train.py
$(dirname "$0")/train.py
Thx for your job!
But I have a question , After I have trained the network how can i get the visualization?
somebody know how to use the demo.py?
what is the command?
I can't reproduce what you did with hrnetsmallv1, i got miou 69.1. And i could not reproduce hrnetsmallv2 with your code. i got miou 70.8. Is there any differnet with you?
Is is possible to use PointRend with HardNet?
Traceback (most recent call last):
File "tools/train.py", line 272, in
trainer.train()
File "tools/train.py", line 163, in train
losses.backward()
File "/home/xjl/.conda/envs/env_torch/lib/python3.6/site-packages/torch/tensor.py", line 107, in backward
torch.autograd.backward(self, gradient, retain_graph, create_graph)
File "/home/xjl/.conda/envs/env_torch/lib/python3.6/site-packages/torch/autograd/init.py", line 93, in backward
allow_unreachable=True) # allow_unreachable flag
RuntimeError: cuda runtime error (11) : invalid argument at /pytorch/aten/src/THC/THCGeneral.cpp:383
the problem occurred during training.how to slove it?thank you
I tried to train DeepLabv3_plus model for pascal_voc by using
CUDA_VISIBLE_DEVICES=0 python-u tools/train.py --config-file configs/pascak_voc_deeplabv3_plus.yaml
due to my configuration problem, modifying the batch size to 4, others is similar to you. But , to 50 epoch , I only achieve 62 mIoU.
Similarly, I tried to train DeepLabv3_plus model for cityscape by using
CUDA_VISIBLE_DEVICES=0 python -u tools/train.py --config-file configs/cityscapes_deeplabv3_plus.yaml
I also achieve about 65 mIoU. So , I want to know whether the pre-training model is used?
It worked by single GPU traing. But it failed no matter how many GPUs I appointed when I tried distributed training.
Line 109 in 4bc605e
self.model.cuda()
It works when I change this line.
The following part is the message of error I met with the former code:
(faceparsing) mjq@amax:~/SegmenTron$ CUDA_VISIBLE_DEVICES=0,7 ./tools/dist_train.sh ${CONFIG_FILE} configs/pascal_voc_deeplabv3_plus.yaml ${GPU_NUM} 2
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.
2020-06-06 02:21:55,815 Segmentron INFO: Using 2 GPUs
2020-06-06 02:21:55,816 Segmentron INFO: Namespace(config_file='configs/pascal_voc_deeplabv3_plus.yaml', device='cuda', distributed=True, input_img='tools/demo_vis.png', local_rank=0, log_iter=10, no_cuda=False, num_gpus=2, opts=[], resume=None, skip_val=False, val_epoch=1)
2020-06-06 02:21:55,816 Segmentron INFO: {
"SEED": 1024,
"TIME_STAMP": "2020-06-06-02-21",
"ROOT_PATH": "/data1/mjq/SegmenTron",
"PHASE": "train",
"DATASET": {
"NAME": "pascal_voc",
"MEAN": [
0.5,
0.5,
0.5
],
"STD": [
0.5,
0.5,
0.5
],
"IGNORE_INDEX": -1,
"WORKERS": 4,
"MODE": "val"
},
"AUG": {
"MIRROR": true,
"BLUR_PROB": 0.0,
"BLUR_RADIUS": 0.0,
"COLOR_JITTER": null
},
"TRAIN": {
"EPOCHS": 50,
"BATCH_SIZE": 4,
"CROP_SIZE": 480,
"BASE_SIZE": 520,
"MODEL_SAVE_DIR": "runs/checkpoints/",
"LOG_SAVE_DIR": "runs/logs/",
"PRETRAINED_MODEL_PATH": "",
"BACKBONE_PRETRAINED": true,
"BACKBONE_PRETRAINED_PATH": "",
"RESUME_MODEL_PATH": "",
"SYNC_BATCH_NORM": true,
"SNAPSHOT_EPOCH": 10
},
"SOLVER": {
"LR": 0.0001,
"OPTIMIZER": "sgd",
"EPSILON": 1e-08,
"MOMENTUM": 0.9,
"WEIGHT_DECAY": 0.0001,
"DECODER_LR_FACTOR": 10.0,
"LR_SCHEDULER": "poly",
"POLY": {
"POWER": 0.9
},
"STEP": {
"GAMMA": 0.1,
"DECAY_EPOCH": [
10,
20
]
},
"WARMUP": {
"EPOCHS": 0.0,
"FACTOR": 0.3333333333333333,
"METHOD": "linear"
},
"OHEM": false,
"AUX": false,
"AUX_WEIGHT": 0.4,
"LOSS_NAME": ""
},
"TEST": {
"TEST_MODEL_PATH": "",
"BATCH_SIZE": 8,
"CROP_SIZE": null,
"SCALES": [
1.0
],
"FLIP": false
},
"VISUAL": {
"OUTPUT_DIR": "../runs/visual/"
},
"MODEL": {
"MODEL_NAME": "DeepLabV3_Plus",
"BACKBONE": "xception65",
"BACKBONE_SCALE": 1.0,
"MULTI_LOSS_WEIGHT": [
1.0
],
"DEFAULT_GROUP_NUMBER": 32,
"DEFAULT_EPSILON": 1e-05,
"BN_TYPE": "BN",
"BN_EPS_FOR_ENCODER": 0.001,
"BN_EPS_FOR_DECODER": null,
"OUTPUT_STRIDE": 16,
"BN_MOMENTUM": null,
"DEEPLABV3_PLUS": {
"USE_ASPP": true,
"ENABLE_DECODER": true,
"ASPP_WITH_SEP_CONV": true,
"DECODER_USE_SEP_CONV": true
},
"CCNET": {
"RECURRENCE": 2
}
}
}
Found 1464 images in the folder datasets/voc/VOC2012
Found 1464 images in the folder datasets/voc/VOC2012
Found 1449 images in the folder datasets/voc/VOC2012
Found 1449 images in the folder datasets/voc/VOC2012
2020-06-06 02:21:56,181 Segmentron INFO: load backbone pretrained model from url..
2020-06-06 02:21:56,480 Segmentron INFO:
Traceback (most recent call last):
File "./tools/train.py", line 223, in
trainer = Trainer(args)
File "./tools/train.py", line 112, in init
find_unused_parameters=True)
File "/data1/mjq/anaconda3/envs/faceparsing/lib/python3.6/site-packages/torch/nn/parallel/distributed.py", line 232, in init
).format(device_ids, output_device, {p.device for p in module.parameters()})
AssertionError: DistributedDataParallel device_ids and output_device arguments only work with single-device CUDA modules, but got device_ids [1], output_device 1, and module parameters {device(type='cuda', index=1), device(type='cpu')}.
2020-06-06 02:21:57,748 Segmentron INFO: DeepLabV3Plus flops: 413.257G input shape is [3, 1024, 2048], params: 41.055M
2020-06-06 02:21:57,776 Segmentron INFO: SyncBatchNorm is effective!
2020-06-06 02:21:57,776 Segmentron INFO: Set bn custom eps for bn in encoder: 0.001
Traceback (most recent call last):
File "./tools/train.py", line 223, in
trainer = Trainer(args)
File "./tools/train.py", line 112, in init
find_unused_parameters=True)
File "/data1/mjq/anaconda3/envs/faceparsing/lib/python3.6/site-packages/torch/nn/parallel/distributed.py", line 232, in init
).format(device_ids, output_device, {p.device for p in module.parameters()})
AssertionError: DistributedDataParallel device_ids and output_device arguments only work with single-device CUDA modules, but got device_ids [0], output_device 0, and module parameters {device(type='cuda', index=0), device(type='cpu')}.
Traceback (most recent call last):
File "/data1/mjq/anaconda3/envs/faceparsing/lib/python3.6/runpy.py", line 193, in _run_module_as_main
"main", mod_spec)
File "/data1/mjq/anaconda3/envs/faceparsing/lib/python3.6/runpy.py", line 85, in _run_code
exec(code, run_globals)
File "/data1/mjq/anaconda3/envs/faceparsing/lib/python3.6/site-packages/torch/distributed/launch.py", line 263, in
main()
File "/data1/mjq/anaconda3/envs/faceparsing/lib/python3.6/site-packages/torch/distributed/launch.py", line 259, in main
cmd=cmd)
subprocess.CalledProcessError: Command '['/data1/mjq/anaconda3/envs/faceparsing/bin/python', '-u', './tools/train.py', '--local_rank=1', '--config-file', 'configs/pascal_voc_deeplabv3_plus.yaml']' returned non-zero exit status 1.
Thanks for your attention! @LikeLy-Journey
Without any change, I train with 8 gpus but only get 59.0 mIoU on cityscapes. But using your trained fast_scnn_segmentron.pth can reproduce 68.9
Did you change something later about the config yaml?
Thanks for your nice work!
Can you modify the code so that it can read the data set correctly under the windows system?
hi ,it is a good job though only 2 stars , please continue to maintain the job
and give the custom dataload and config. haha ,thansks
@LikeLy-Journey thanks for your greak work ! when i run tools/eval.py i ran into this problem,and i am use the config of the pascal_voc_deeplabv3_plus.yaml,and i change the BACKBONE: "xception65" to BACKBONE: "resnet50" because of my low gpu
memory,so what should i do to fix the problem? thanks
Hi.Thank u for your code. When I run the code, It is wrong,but I don't know why.Thank you for your help.:
2020-02-29 09:16:37,190 Segmentron INFO:
2020-02-29 09:16:51,159 Segmentron INFO: DeepLabV3Plus flops: 458.505G input shape is [3, 1024, 2048], params: 47.737M
2020-02-29 09:16:51,161 Segmentron INFO: Not use SyncBatchNorm!
2020-02-29 09:16:51,163 Segmentron INFO: Start training, Total Epochs: 400 = Total Iterations 74000
2020-02-29 09:16:57,141 Segmentron INFO: DeepLabV3Plus flops: 458.505G input shape is [3, 1024, 2048], params: 47.737M
2020-02-29 09:16:57,143 Segmentron INFO: Not use SyncBatchNorm!
2020-02-29 09:16:57,145 Segmentron INFO: Start training, Total Epochs: 400 = Total Iterations 74000
2020-02-29 09:17:02,040 Segmentron INFO: DeepLabV3Plus flops: 458.505G input shape is [3, 1024, 2048], params: 47.737M
2020-02-29 09:17:02,042 Segmentron INFO: Not use SyncBatchNorm!
2020-02-29 09:17:02,044 Segmentron INFO: Start training, Total Epochs: 400 = Total Iterations 74000
2020-02-29 09:17:02,214 Segmentron INFO: DeepLabV3Plus flops: 458.505G input shape is [3, 1024, 2048], params: 47.737M
2020-02-29 09:17:02,216 Segmentron INFO: Not use SyncBatchNorm!
2020-02-29 09:17:02,218 Segmentron INFO: Start training, Total Epochs: 400 = Total Iterations 74000
Traceback (most recent call last):
File "./tools/train.py", line 221, in
trainer.train()
File "./tools/train.py", line 133, in train
outputs = self.model(images)
File "/usr/local/lib/python3.5/dist-packages/torch/nn/modules/module.py", line 532, in call
result = self.forward(*input, **kwargs)
File "/root/seg/segmentron/models/deeplabv3_plus.py", line 38, in forward
x = self.head(c4, c1)
File "/usr/local/lib/python3.5/dist-packages/torch/nn/modules/module.py", line 532, in call
result = self.forward(*input, **kwargs)
File "/root/seg/segmentron/models/deeplabv3_plus.py", line 69, in forward
x = self.aspp(x)
File "/usr/local/lib/python3.5/dist-packages/torch/nn/modules/module.py", line 532, in call
result = self.forward(*input, **kwargs)
File "/root/seg/segmentron/modules/module.py", line 66, in forward
x0 = self.aspp0(x)
File "/usr/local/lib/python3.5/dist-packages/torch/nn/modules/module.py", line 532, in call
result = self.forward(*input, **kwargs)
File "/usr/local/lib/python3.5/dist-packages/torch/nn/modules/container.py", line 100, in forward
input = module(input)
File "/usr/local/lib/python3.5/dist-packages/torch/nn/modules/module.py", line 532, in call
result = self.forward(*input, **kwargs)
File "/usr/local/lib/python3.5/dist-packages/torch/nn/modules/conv.py", line 345, in forward
return self.conv2d_forward(input, self.weight)
File "/usr/local/lib/python3.5/dist-packages/torch/nn/modules/conv.py", line 342, in conv2d_forward
self.padding, self.dilation, self.groups)
File "/usr/local/lib/python3.5/dist-packages/torch/utils/data/_utils/signal_handling.py", line 66, in handler
_error_if_any_worker_fails()
RuntimeError: DataLoader worker (pid 8737) is killed by signal: Killed.
Traceback (most recent call last):
File "/usr/lib/python3.5/runpy.py", line 184, in _run_module_as_main
"main", mod_spec)
File "/usr/lib/python3.5/runpy.py", line 85, in _run_code
exec(code, run_globals)
File "/usr/local/lib/python3.5/dist-packages/torch/distributed/launch.py", line 263, in
main()
File "/usr/local/lib/python3.5/dist-packages/torch/distributed/launch.py", line 259, in main
cmd=cmd)
subprocess.CalledProcessError: Command '['/usr/bin/python', '-u', './tools/train.py', '--local_rank=3', '--config-file', 'configs/cityscapes_deeplabv3_plus_resnet.yaml']' returned non-zero exit status -9
Very nice repo! Hope for more models (like refinenet) and more datasets (like SUNRGBD and NYUDV2).
A problem is: my cuda memory is ok if I train without loading a model. However, when training is stopped, and I set args.resume to the current checkpoint for continue training, cuda will be out of memory.
The config file I use is cityscapes_deeplabv3_plus.yaml. I don't know why.
@LikeLy-Journey When using HRNet_seg I am getting this error, it is owed to loading_ditc.
~/Documents/proy1/seg/models/archs/backbones/build.py in get_segmentation_backbone(backbone, norm_layer)
60 """
61 model = BACKBONE_REGISTRY.get(backbone)(norm_layer)
---> 62 load_backbone_pretrained(model, backbone)
63 return model
64
~/Documents/proy1/seg/models/archs/backbones/build.py in load_backbone_pretrained(model, backbone)
51
52 msg = model.load_state_dict(torch.load(download(model_urls[backbone],
---> 53 path=os.path.join(torch.hub._get_torch_home(), 'checkpoints'))), strict=False)
54 logging.info(msg)
55
~/anaconda3/envs/seg/lib/python3.7/site-packages/torch/nn/modules/module.py in load_state_dict(self, state_dict, strict)
853 if len(error_msgs) > 0:
854 raise RuntimeError('Error(s) in loading state_dict for {}:\n\t{}'.format(
--> 855 self.__class__.__name__, "\n\t".join(error_msgs)))
856 return _IncompatibleKeys(missing_keys, unexpected_keys)
857
RuntimeError: Error(s) in loading state_dict for HighResolutionNet:
size mismatch for transition1.0.0.weight: copying a param with shape torch.Size([16, 128, 3, 3]) from checkpoint, the shape in current model is torch.Size([32, 128, 3, 3]).
size mismatch for transition1.0.1.weight: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for transition1.0.1.bias: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for transition1.0.1.running_mean: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for transition1.0.1.running_var: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for transition1.1.0.0.weight: copying a param with shape torch.Size([32, 128, 3, 3]) from checkpoint, the shape in current model is torch.Size([64, 128, 3, 3]).
size mismatch for transition1.1.0.1.weight: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for transition1.1.0.1.bias: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for transition1.1.0.1.running_mean: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for transition1.1.0.1.running_var: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for stage2.0.branches.0.0.conv1.weight: copying a param with shape torch.Size([16, 16, 3, 3]) from checkpoint, the shape in current model is torch.Size([32, 32, 3, 3]).
size mismatch for stage2.0.branches.0.0.bn1.weight: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage2.0.branches.0.0.bn1.bias: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage2.0.branches.0.0.bn1.running_mean: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage2.0.branches.0.0.bn1.running_var: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage2.0.branches.0.0.conv2.weight: copying a param with shape torch.Size([16, 16, 3, 3]) from checkpoint, the shape in current model is torch.Size([32, 32, 3, 3]).
size mismatch for stage2.0.branches.0.0.bn2.weight: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage2.0.branches.0.0.bn2.bias: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage2.0.branches.0.0.bn2.running_mean: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage2.0.branches.0.0.bn2.running_var: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage2.0.branches.0.1.conv1.weight: copying a param with shape torch.Size([16, 16, 3, 3]) from checkpoint, the shape in current model is torch.Size([32, 32, 3, 3]).
size mismatch for stage2.0.branches.0.1.bn1.weight: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage2.0.branches.0.1.bn1.bias: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage2.0.branches.0.1.bn1.running_mean: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage2.0.branches.0.1.bn1.running_var: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage2.0.branches.0.1.conv2.weight: copying a param with shape torch.Size([16, 16, 3, 3]) from checkpoint, the shape in current model is torch.Size([32, 32, 3, 3]).
size mismatch for stage2.0.branches.0.1.bn2.weight: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage2.0.branches.0.1.bn2.bias: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage2.0.branches.0.1.bn2.running_mean: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage2.0.branches.0.1.bn2.running_var: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage2.0.branches.1.0.conv1.weight: copying a param with shape torch.Size([32, 32, 3, 3]) from checkpoint, the shape in current model is torch.Size([64, 64, 3, 3]).
size mismatch for stage2.0.branches.1.0.bn1.weight: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for stage2.0.branches.1.0.bn1.bias: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for stage2.0.branches.1.0.bn1.running_mean: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for stage2.0.branches.1.0.bn1.running_var: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for stage2.0.branches.1.0.conv2.weight: copying a param with shape torch.Size([32, 32, 3, 3]) from checkpoint, the shape in current model is torch.Size([64, 64, 3, 3]).
size mismatch for stage2.0.branches.1.0.bn2.weight: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for stage2.0.branches.1.0.bn2.bias: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for stage2.0.branches.1.0.bn2.running_mean: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for stage2.0.branches.1.0.bn2.running_var: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for stage2.0.branches.1.1.conv1.weight: copying a param with shape torch.Size([32, 32, 3, 3]) from checkpoint, the shape in current model is torch.Size([64, 64, 3, 3]).
size mismatch for stage2.0.branches.1.1.bn1.weight: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for stage2.0.branches.1.1.bn1.bias: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for stage2.0.branches.1.1.bn1.running_mean: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for stage2.0.branches.1.1.bn1.running_var: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for stage2.0.branches.1.1.conv2.weight: copying a param with shape torch.Size([32, 32, 3, 3]) from checkpoint, the shape in current model is torch.Size([64, 64, 3, 3]).
size mismatch for stage2.0.branches.1.1.bn2.weight: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for stage2.0.branches.1.1.bn2.bias: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for stage2.0.branches.1.1.bn2.running_mean: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for stage2.0.branches.1.1.bn2.running_var: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for stage2.0.fuse_layers.0.1.0.weight: copying a param with shape torch.Size([16, 32, 1, 1]) from checkpoint, the shape in current model is torch.Size([32, 64, 1, 1]).
size mismatch for stage2.0.fuse_layers.0.1.1.weight: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage2.0.fuse_layers.0.1.1.bias: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage2.0.fuse_layers.0.1.1.running_mean: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage2.0.fuse_layers.0.1.1.running_var: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage2.0.fuse_layers.1.0.0.0.weight: copying a param with shape torch.Size([32, 16, 3, 3]) from checkpoint, the shape in current model is torch.Size([64, 32, 3, 3]).
size mismatch for stage2.0.fuse_layers.1.0.0.1.weight: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for stage2.0.fuse_layers.1.0.0.1.bias: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for stage2.0.fuse_layers.1.0.0.1.running_mean: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for stage2.0.fuse_layers.1.0.0.1.running_var: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for transition2.2.0.0.weight: copying a param with shape torch.Size([64, 32, 3, 3]) from checkpoint, the shape in current model is torch.Size([128, 64, 3, 3]).
size mismatch for transition2.2.0.1.weight: copying a param with shape torch.Size([64]) from checkpoint, the shape in current model is torch.Size([128]).
size mismatch for transition2.2.0.1.bias: copying a param with shape torch.Size([64]) from checkpoint, the shape in current model is torch.Size([128]).
size mismatch for transition2.2.0.1.running_mean: copying a param with shape torch.Size([64]) from checkpoint, the shape in current model is torch.Size([128]).
size mismatch for transition2.2.0.1.running_var: copying a param with shape torch.Size([64]) from checkpoint, the shape in current model is torch.Size([128]).
size mismatch for stage3.0.branches.0.0.conv1.weight: copying a param with shape torch.Size([16, 16, 3, 3]) from checkpoint, the shape in current model is torch.Size([32, 32, 3, 3]).
size mismatch for stage3.0.branches.0.0.bn1.weight: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage3.0.branches.0.0.bn1.bias: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage3.0.branches.0.0.bn1.running_mean: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage3.0.branches.0.0.bn1.running_var: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage3.0.branches.0.0.conv2.weight: copying a param with shape torch.Size([16, 16, 3, 3]) from checkpoint, the shape in current model is torch.Size([32, 32, 3, 3]).
size mismatch for stage3.0.branches.0.0.bn2.weight: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage3.0.branches.0.0.bn2.bias: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage3.0.branches.0.0.bn2.running_mean: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage3.0.branches.0.0.bn2.running_var: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage3.0.branches.0.1.conv1.weight: copying a param with shape torch.Size([16, 16, 3, 3]) from checkpoint, the shape in current model is torch.Size([32, 32, 3, 3]).
size mismatch for stage3.0.branches.0.1.bn1.weight: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage3.0.branches.0.1.bn1.bias: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage3.0.branches.0.1.bn1.running_mean: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage3.0.branches.0.1.bn1.running_var: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage3.0.branches.0.1.conv2.weight: copying a param with shape torch.Size([16, 16, 3, 3]) from checkpoint, the shape in current model is torch.Size([32, 32, 3, 3]).
size mismatch for stage3.0.branches.0.1.bn2.weight: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage3.0.branches.0.1.bn2.bias: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage3.0.branches.0.1.bn2.running_mean: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage3.0.branches.0.1.bn2.running_var: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage3.0.branches.1.0.conv1.weight: copying a param with shape torch.Size([32, 32, 3, 3]) from checkpoint, the shape in current model is torch.Size([64, 64, 3, 3]).
size mismatch for stage3.0.branches.1.0.bn1.weight: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for stage3.0.branches.1.0.bn1.bias: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for stage3.0.branches.1.0.bn1.running_mean: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for stage3.0.branches.1.0.bn1.running_var: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for stage3.0.branches.1.0.conv2.weight: copying a param with shape torch.Size([32, 32, 3, 3]) from checkpoint, the shape in current model is torch.Size([64, 64, 3, 3]).
size mismatch for stage3.0.branches.1.0.bn2.weight: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for stage3.0.branches.1.0.bn2.bias: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for stage3.0.branches.1.0.bn2.running_mean: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for stage3.0.branches.1.0.bn2.running_var: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for stage3.0.branches.1.1.conv1.weight: copying a param with shape torch.Size([32, 32, 3, 3]) from checkpoint, the shape in current model is torch.Size([64, 64, 3, 3]).
size mismatch for stage3.0.branches.1.1.bn1.weight: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for stage3.0.branches.1.1.bn1.bias: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for stage3.0.branches.1.1.bn1.running_mean: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for stage3.0.branches.1.1.bn1.running_var: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for stage3.0.branches.1.1.conv2.weight: copying a param with shape torch.Size([32, 32, 3, 3]) from checkpoint, the shape in current model is torch.Size([64, 64, 3, 3]).
size mismatch for stage3.0.branches.1.1.bn2.weight: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for stage3.0.branches.1.1.bn2.bias: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for stage3.0.branches.1.1.bn2.running_mean: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for stage3.0.branches.1.1.bn2.running_var: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for stage3.0.branches.2.0.conv1.weight: copying a param with shape torch.Size([64, 64, 3, 3]) from checkpoint, the shape in current model is torch.Size([128, 128, 3, 3]).
size mismatch for stage3.0.branches.2.0.bn1.weight: copying a param with shape torch.Size([64]) from checkpoint, the shape in current model is torch.Size([128]).
size mismatch for stage3.0.branches.2.0.bn1.bias: copying a param with shape torch.Size([64]) from checkpoint, the shape in current model is torch.Size([128]).
size mismatch for stage3.0.branches.2.0.bn1.running_mean: copying a param with shape torch.Size([64]) from checkpoint, the shape in current model is torch.Size([128]).
size mismatch for stage3.0.branches.2.0.bn1.running_var: copying a param with shape torch.Size([64]) from checkpoint, the shape in current model is torch.Size([128]).
size mismatch for stage3.0.branches.2.0.conv2.weight: copying a param with shape torch.Size([64, 64, 3, 3]) from checkpoint, the shape in current model is torch.Size([128, 128, 3, 3]).
size mismatch for stage3.0.branches.2.0.bn2.weight: copying a param with shape torch.Size([64]) from checkpoint, the shape in current model is torch.Size([128]).
size mismatch for stage3.0.branches.2.0.bn2.bias: copying a param with shape torch.Size([64]) from checkpoint, the shape in current model is torch.Size([128]).
size mismatch for stage3.0.branches.2.0.bn2.running_mean: copying a param with shape torch.Size([64]) from checkpoint, the shape in current model is torch.Size([128]).
size mismatch for stage3.0.branches.2.0.bn2.running_var: copying a param with shape torch.Size([64]) from checkpoint, the shape in current model is torch.Size([128]).
size mismatch for stage3.0.branches.2.1.conv1.weight: copying a param with shape torch.Size([64, 64, 3, 3]) from checkpoint, the shape in current model is torch.Size([128, 128, 3, 3]).
size mismatch for stage3.0.branches.2.1.bn1.weight: copying a param with shape torch.Size([64]) from checkpoint, the shape in current model is torch.Size([128]).
size mismatch for stage3.0.branches.2.1.bn1.bias: copying a param with shape torch.Size([64]) from checkpoint, the shape in current model is torch.Size([128]).
size mismatch for stage3.0.branches.2.1.bn1.running_mean: copying a param with shape torch.Size([64]) from checkpoint, the shape in current model is torch.Size([128]).
size mismatch for stage3.0.branches.2.1.bn1.running_var: copying a param with shape torch.Size([64]) from checkpoint, the shape in current model is torch.Size([128]).
size mismatch for stage3.0.branches.2.1.conv2.weight: copying a param with shape torch.Size([64, 64, 3, 3]) from checkpoint, the shape in current model is torch.Size([128, 128, 3, 3]).
size mismatch for stage3.0.branches.2.1.bn2.weight: copying a param with shape torch.Size([64]) from checkpoint, the shape in current model is torch.Size([128]).
size mismatch for stage3.0.branches.2.1.bn2.bias: copying a param with shape torch.Size([64]) from checkpoint, the shape in current model is torch.Size([128]).
size mismatch for stage3.0.branches.2.1.bn2.running_mean: copying a param with shape torch.Size([64]) from checkpoint, the shape in current model is torch.Size([128]).
size mismatch for stage3.0.branches.2.1.bn2.running_var: copying a param with shape torch.Size([64]) from checkpoint, the shape in current model is torch.Size([128]).
size mismatch for stage3.0.fuse_layers.0.1.0.weight: copying a param with shape torch.Size([16, 32, 1, 1]) from checkpoint, the shape in current model is torch.Size([32, 64, 1, 1]).
size mismatch for stage3.0.fuse_layers.0.1.1.weight: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage3.0.fuse_layers.0.1.1.bias: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage3.0.fuse_layers.0.1.1.running_mean: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage3.0.fuse_layers.0.1.1.running_var: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage3.0.fuse_layers.0.2.0.weight: copying a param with shape torch.Size([16, 64, 1, 1]) from checkpoint, the shape in current model is torch.Size([32, 128, 1, 1]).
size mismatch for stage3.0.fuse_layers.0.2.1.weight: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage3.0.fuse_layers.0.2.1.bias: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage3.0.fuse_layers.0.2.1.running_mean: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage3.0.fuse_layers.0.2.1.running_var: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage3.0.fuse_layers.1.0.0.0.weight: copying a param with shape torch.Size([32, 16, 3, 3]) from checkpoint, the shape in current model is torch.Size([64, 32, 3, 3]).
size mismatch for stage3.0.fuse_layers.1.0.0.1.weight: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for stage3.0.fuse_layers.1.0.0.1.bias: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for stage3.0.fuse_layers.1.0.0.1.running_mean: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for stage3.0.fuse_layers.1.0.0.1.running_var: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for stage3.0.fuse_layers.1.2.0.weight: copying a param with shape torch.Size([32, 64, 1, 1]) from checkpoint, the shape in current model is torch.Size([64, 128, 1, 1]).
size mismatch for stage3.0.fuse_layers.1.2.1.weight: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for stage3.0.fuse_layers.1.2.1.bias: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for stage3.0.fuse_layers.1.2.1.running_mean: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for stage3.0.fuse_layers.1.2.1.running_var: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for stage3.0.fuse_layers.2.0.0.0.weight: copying a param with shape torch.Size([16, 16, 3, 3]) from checkpoint, the shape in current model is torch.Size([32, 32, 3, 3]).
size mismatch for stage3.0.fuse_layers.2.0.0.1.weight: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage3.0.fuse_layers.2.0.0.1.bias: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage3.0.fuse_layers.2.0.0.1.running_mean: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage3.0.fuse_layers.2.0.0.1.running_var: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage3.0.fuse_layers.2.0.1.0.weight: copying a param with shape torch.Size([64, 16, 3, 3]) from checkpoint, the shape in current model is torch.Size([128, 32, 3, 3]).
size mismatch for stage3.0.fuse_layers.2.0.1.1.weight: copying a param with shape torch.Size([64]) from checkpoint, the shape in current model is torch.Size([128]).
size mismatch for stage3.0.fuse_layers.2.0.1.1.bias: copying a param with shape torch.Size([64]) from checkpoint, the shape in current model is torch.Size([128]).
size mismatch for stage3.0.fuse_layers.2.0.1.1.running_mean: copying a param with shape torch.Size([64]) from checkpoint, the shape in current model is torch.Size([128]).
size mismatch for stage3.0.fuse_layers.2.0.1.1.running_var: copying a param with shape torch.Size([64]) from checkpoint, the shape in current model is torch.Size([128]).
size mismatch for stage3.0.fuse_layers.2.1.0.0.weight: copying a param with shape torch.Size([64, 32, 3, 3]) from checkpoint, the shape in current model is torch.Size([128, 64, 3, 3]).
size mismatch for stage3.0.fuse_layers.2.1.0.1.weight: copying a param with shape torch.Size([64]) from checkpoint, the shape in current model is torch.Size([128]).
size mismatch for stage3.0.fuse_layers.2.1.0.1.bias: copying a param with shape torch.Size([64]) from checkpoint, the shape in current model is torch.Size([128]).
size mismatch for stage3.0.fuse_layers.2.1.0.1.running_mean: copying a param with shape torch.Size([64]) from checkpoint, the shape in current model is torch.Size([128]).
size mismatch for stage3.0.fuse_layers.2.1.0.1.running_var: copying a param with shape torch.Size([64]) from checkpoint, the shape in current model is torch.Size([128]).
size mismatch for transition3.3.0.0.weight: copying a param with shape torch.Size([128, 64, 3, 3]) from checkpoint, the shape in current model is torch.Size([256, 128, 3, 3]).
size mismatch for transition3.3.0.1.weight: copying a param with shape torch.Size([128]) from checkpoint, the shape in current model is torch.Size([256]).
size mismatch for transition3.3.0.1.bias: copying a param with shape torch.Size([128]) from checkpoint, the shape in current model is torch.Size([256]).
size mismatch for transition3.3.0.1.running_mean: copying a param with shape torch.Size([128]) from checkpoint, the shape in current model is torch.Size([256]).
size mismatch for transition3.3.0.1.running_var: copying a param with shape torch.Size([128]) from checkpoint, the shape in current model is torch.Size([256]).
size mismatch for stage4.0.branches.0.0.conv1.weight: copying a param with shape torch.Size([16, 16, 3, 3]) from checkpoint, the shape in current model is torch.Size([32, 32, 3, 3]).
size mismatch for stage4.0.branches.0.0.bn1.weight: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage4.0.branches.0.0.bn1.bias: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage4.0.branches.0.0.bn1.running_mean: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage4.0.branches.0.0.bn1.running_var: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage4.0.branches.0.0.conv2.weight: copying a param with shape torch.Size([16, 16, 3, 3]) from checkpoint, the shape in current model is torch.Size([32, 32, 3, 3]).
size mismatch for stage4.0.branches.0.0.bn2.weight: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage4.0.branches.0.0.bn2.bias: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage4.0.branches.0.0.bn2.running_mean: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage4.0.branches.0.0.bn2.running_var: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage4.0.branches.0.1.conv1.weight: copying a param with shape torch.Size([16, 16, 3, 3]) from checkpoint, the shape in current model is torch.Size([32, 32, 3, 3]).
size mismatch for stage4.0.branches.0.1.bn1.weight: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage4.0.branches.0.1.bn1.bias: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage4.0.branches.0.1.bn1.running_mean: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage4.0.branches.0.1.bn1.running_var: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage4.0.branches.0.1.conv2.weight: copying a param with shape torch.Size([16, 16, 3, 3]) from checkpoint, the shape in current model is torch.Size([32, 32, 3, 3]).
size mismatch for stage4.0.branches.0.1.bn2.weight: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage4.0.branches.0.1.bn2.bias: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage4.0.branches.0.1.bn2.running_mean: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage4.0.branches.0.1.bn2.running_var: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage4.0.branches.1.0.conv1.weight: copying a param with shape torch.Size([32, 32, 3, 3]) from checkpoint, the shape in current model is torch.Size([64, 64, 3, 3]).
size mismatch for stage4.0.branches.1.0.bn1.weight: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for stage4.0.branches.1.0.bn1.bias: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for stage4.0.branches.1.0.bn1.running_mean: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for stage4.0.branches.1.0.bn1.running_var: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for stage4.0.branches.1.0.conv2.weight: copying a param with shape torch.Size([32, 32, 3, 3]) from checkpoint, the shape in current model is torch.Size([64, 64, 3, 3]).
size mismatch for stage4.0.branches.1.0.bn2.weight: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for stage4.0.branches.1.0.bn2.bias: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for stage4.0.branches.1.0.bn2.running_mean: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for stage4.0.branches.1.0.bn2.running_var: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for stage4.0.branches.1.1.conv1.weight: copying a param with shape torch.Size([32, 32, 3, 3]) from checkpoint, the shape in current model is torch.Size([64, 64, 3, 3]).
size mismatch for stage4.0.branches.1.1.bn1.weight: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for stage4.0.branches.1.1.bn1.bias: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for stage4.0.branches.1.1.bn1.running_mean: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for stage4.0.branches.1.1.bn1.running_var: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for stage4.0.branches.1.1.conv2.weight: copying a param with shape torch.Size([32, 32, 3, 3]) from checkpoint, the shape in current model is torch.Size([64, 64, 3, 3]).
size mismatch for stage4.0.branches.1.1.bn2.weight: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for stage4.0.branches.1.1.bn2.bias: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for stage4.0.branches.1.1.bn2.running_mean: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for stage4.0.branches.1.1.bn2.running_var: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for stage4.0.branches.2.0.conv1.weight: copying a param with shape torch.Size([64, 64, 3, 3]) from checkpoint, the shape in current model is torch.Size([128, 128, 3, 3]).
size mismatch for stage4.0.branches.2.0.bn1.weight: copying a param with shape torch.Size([64]) from checkpoint, the shape in current model is torch.Size([128]).
size mismatch for stage4.0.branches.2.0.bn1.bias: copying a param with shape torch.Size([64]) from checkpoint, the shape in current model is torch.Size([128]).
size mismatch for stage4.0.branches.2.0.bn1.running_mean: copying a param with shape torch.Size([64]) from checkpoint, the shape in current model is torch.Size([128]).
size mismatch for stage4.0.branches.2.0.bn1.running_var: copying a param with shape torch.Size([64]) from checkpoint, the shape in current model is torch.Size([128]).
size mismatch for stage4.0.branches.2.0.conv2.weight: copying a param with shape torch.Size([64, 64, 3, 3]) from checkpoint, the shape in current model is torch.Size([128, 128, 3, 3]).
size mismatch for stage4.0.branches.2.0.bn2.weight: copying a param with shape torch.Size([64]) from checkpoint, the shape in current model is torch.Size([128]).
size mismatch for stage4.0.branches.2.0.bn2.bias: copying a param with shape torch.Size([64]) from checkpoint, the shape in current model is torch.Size([128]).
size mismatch for stage4.0.branches.2.0.bn2.running_mean: copying a param with shape torch.Size([64]) from checkpoint, the shape in current model is torch.Size([128]).
size mismatch for stage4.0.branches.2.0.bn2.running_var: copying a param with shape torch.Size([64]) from checkpoint, the shape in current model is torch.Size([128]).
size mismatch for stage4.0.branches.2.1.conv1.weight: copying a param with shape torch.Size([64, 64, 3, 3]) from checkpoint, the shape in current model is torch.Size([128, 128, 3, 3]).
size mismatch for stage4.0.branches.2.1.bn1.weight: copying a param with shape torch.Size([64]) from checkpoint, the shape in current model is torch.Size([128]).
size mismatch for stage4.0.branches.2.1.bn1.bias: copying a param with shape torch.Size([64]) from checkpoint, the shape in current model is torch.Size([128]).
size mismatch for stage4.0.branches.2.1.bn1.running_mean: copying a param with shape torch.Size([64]) from checkpoint, the shape in current model is torch.Size([128]).
size mismatch for stage4.0.branches.2.1.bn1.running_var: copying a param with shape torch.Size([64]) from checkpoint, the shape in current model is torch.Size([128]).
size mismatch for stage4.0.branches.2.1.conv2.weight: copying a param with shape torch.Size([64, 64, 3, 3]) from checkpoint, the shape in current model is torch.Size([128, 128, 3, 3]).
size mismatch for stage4.0.branches.2.1.bn2.weight: copying a param with shape torch.Size([64]) from checkpoint, the shape in current model is torch.Size([128]).
size mismatch for stage4.0.branches.2.1.bn2.bias: copying a param with shape torch.Size([64]) from checkpoint, the shape in current model is torch.Size([128]).
size mismatch for stage4.0.branches.2.1.bn2.running_mean: copying a param with shape torch.Size([64]) from checkpoint, the shape in current model is torch.Size([128]).
size mismatch for stage4.0.branches.2.1.bn2.running_var: copying a param with shape torch.Size([64]) from checkpoint, the shape in current model is torch.Size([128]).
size mismatch for stage4.0.branches.3.0.conv1.weight: copying a param with shape torch.Size([128, 128, 3, 3]) from checkpoint, the shape in current model is torch.Size([256, 256, 3, 3]).
size mismatch for stage4.0.branches.3.0.bn1.weight: copying a param with shape torch.Size([128]) from checkpoint, the shape in current model is torch.Size([256]).
size mismatch for stage4.0.branches.3.0.bn1.bias: copying a param with shape torch.Size([128]) from checkpoint, the shape in current model is torch.Size([256]).
size mismatch for stage4.0.branches.3.0.bn1.running_mean: copying a param with shape torch.Size([128]) from checkpoint, the shape in current model is torch.Size([256]).
size mismatch for stage4.0.branches.3.0.bn1.running_var: copying a param with shape torch.Size([128]) from checkpoint, the shape in current model is torch.Size([256]).
size mismatch for stage4.0.branches.3.0.conv2.weight: copying a param with shape torch.Size([128, 128, 3, 3]) from checkpoint, the shape in current model is torch.Size([256, 256, 3, 3]).
size mismatch for stage4.0.branches.3.0.bn2.weight: copying a param with shape torch.Size([128]) from checkpoint, the shape in current model is torch.Size([256]).
size mismatch for stage4.0.branches.3.0.bn2.bias: copying a param with shape torch.Size([128]) from checkpoint, the shape in current model is torch.Size([256]).
size mismatch for stage4.0.branches.3.0.bn2.running_mean: copying a param with shape torch.Size([128]) from checkpoint, the shape in current model is torch.Size([256]).
size mismatch for stage4.0.branches.3.0.bn2.running_var: copying a param with shape torch.Size([128]) from checkpoint, the shape in current model is torch.Size([256]).
size mismatch for stage4.0.branches.3.1.conv1.weight: copying a param with shape torch.Size([128, 128, 3, 3]) from checkpoint, the shape in current model is torch.Size([256, 256, 3, 3]).
size mismatch for stage4.0.branches.3.1.bn1.weight: copying a param with shape torch.Size([128]) from checkpoint, the shape in current model is torch.Size([256]).
size mismatch for stage4.0.branches.3.1.bn1.bias: copying a param with shape torch.Size([128]) from checkpoint, the shape in current model is torch.Size([256]).
size mismatch for stage4.0.branches.3.1.bn1.running_mean: copying a param with shape torch.Size([128]) from checkpoint, the shape in current model is torch.Size([256]).
size mismatch for stage4.0.branches.3.1.bn1.running_var: copying a param with shape torch.Size([128]) from checkpoint, the shape in current model is torch.Size([256]).
size mismatch for stage4.0.branches.3.1.conv2.weight: copying a param with shape torch.Size([128, 128, 3, 3]) from checkpoint, the shape in current model is torch.Size([256, 256, 3, 3]).
size mismatch for stage4.0.branches.3.1.bn2.weight: copying a param with shape torch.Size([128]) from checkpoint, the shape in current model is torch.Size([256]).
size mismatch for stage4.0.branches.3.1.bn2.bias: copying a param with shape torch.Size([128]) from checkpoint, the shape in current model is torch.Size([256]).
size mismatch for stage4.0.branches.3.1.bn2.running_mean: copying a param with shape torch.Size([128]) from checkpoint, the shape in current model is torch.Size([256]).
size mismatch for stage4.0.branches.3.1.bn2.running_var: copying a param with shape torch.Size([128]) from checkpoint, the shape in current model is torch.Size([256]).
size mismatch for stage4.0.fuse_layers.0.1.0.weight: copying a param with shape torch.Size([16, 32, 1, 1]) from checkpoint, the shape in current model is torch.Size([32, 64, 1, 1]).
size mismatch for stage4.0.fuse_layers.0.1.1.weight: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage4.0.fuse_layers.0.1.1.bias: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage4.0.fuse_layers.0.1.1.running_mean: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage4.0.fuse_layers.0.1.1.running_var: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage4.0.fuse_layers.0.2.0.weight: copying a param with shape torch.Size([16, 64, 1, 1]) from checkpoint, the shape in current model is torch.Size([32, 128, 1, 1]).
size mismatch for stage4.0.fuse_layers.0.2.1.weight: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage4.0.fuse_layers.0.2.1.bias: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage4.0.fuse_layers.0.2.1.running_mean: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage4.0.fuse_layers.0.2.1.running_var: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage4.0.fuse_layers.0.3.0.weight: copying a param with shape torch.Size([16, 128, 1, 1]) from checkpoint, the shape in current model is torch.Size([32, 256, 1, 1]).
size mismatch for stage4.0.fuse_layers.0.3.1.weight: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage4.0.fuse_layers.0.3.1.bias: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage4.0.fuse_layers.0.3.1.running_mean: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage4.0.fuse_layers.0.3.1.running_var: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage4.0.fuse_layers.1.0.0.0.weight: copying a param with shape torch.Size([32, 16, 3, 3]) from checkpoint, the shape in current model is torch.Size([64, 32, 3, 3]).
size mismatch for stage4.0.fuse_layers.1.0.0.1.weight: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for stage4.0.fuse_layers.1.0.0.1.bias: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for stage4.0.fuse_layers.1.0.0.1.running_mean: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for stage4.0.fuse_layers.1.0.0.1.running_var: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for stage4.0.fuse_layers.1.2.0.weight: copying a param with shape torch.Size([32, 64, 1, 1]) from checkpoint, the shape in current model is torch.Size([64, 128, 1, 1]).
size mismatch for stage4.0.fuse_layers.1.2.1.weight: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for stage4.0.fuse_layers.1.2.1.bias: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for stage4.0.fuse_layers.1.2.1.running_mean: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for stage4.0.fuse_layers.1.2.1.running_var: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for stage4.0.fuse_layers.1.3.0.weight: copying a param with shape torch.Size([32, 128, 1, 1]) from checkpoint, the shape in current model is torch.Size([64, 256, 1, 1]).
size mismatch for stage4.0.fuse_layers.1.3.1.weight: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for stage4.0.fuse_layers.1.3.1.bias: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for stage4.0.fuse_layers.1.3.1.running_mean: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for stage4.0.fuse_layers.1.3.1.running_var: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for stage4.0.fuse_layers.2.0.0.0.weight: copying a param with shape torch.Size([16, 16, 3, 3]) from checkpoint, the shape in current model is torch.Size([32, 32, 3, 3]).
size mismatch for stage4.0.fuse_layers.2.0.0.1.weight: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage4.0.fuse_layers.2.0.0.1.bias: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage4.0.fuse_layers.2.0.0.1.running_mean: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage4.0.fuse_layers.2.0.0.1.running_var: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage4.0.fuse_layers.2.0.1.0.weight: copying a param with shape torch.Size([64, 16, 3, 3]) from checkpoint, the shape in current model is torch.Size([128, 32, 3, 3]).
size mismatch for stage4.0.fuse_layers.2.0.1.1.weight: copying a param with shape torch.Size([64]) from checkpoint, the shape in current model is torch.Size([128]).
size mismatch for stage4.0.fuse_layers.2.0.1.1.bias: copying a param with shape torch.Size([64]) from checkpoint, the shape in current model is torch.Size([128]).
size mismatch for stage4.0.fuse_layers.2.0.1.1.running_mean: copying a param with shape torch.Size([64]) from checkpoint, the shape in current model is torch.Size([128]).
size mismatch for stage4.0.fuse_layers.2.0.1.1.running_var: copying a param with shape torch.Size([64]) from checkpoint, the shape in current model is torch.Size([128]).
size mismatch for stage4.0.fuse_layers.2.1.0.0.weight: copying a param with shape torch.Size([64, 32, 3, 3]) from checkpoint, the shape in current model is torch.Size([128, 64, 3, 3]).
size mismatch for stage4.0.fuse_layers.2.1.0.1.weight: copying a param with shape torch.Size([64]) from checkpoint, the shape in current model is torch.Size([128]).
size mismatch for stage4.0.fuse_layers.2.1.0.1.bias: copying a param with shape torch.Size([64]) from checkpoint, the shape in current model is torch.Size([128]).
size mismatch for stage4.0.fuse_layers.2.1.0.1.running_mean: copying a param with shape torch.Size([64]) from checkpoint, the shape in current model is torch.Size([128]).
size mismatch for stage4.0.fuse_layers.2.1.0.1.running_var: copying a param with shape torch.Size([64]) from checkpoint, the shape in current model is torch.Size([128]).
size mismatch for stage4.0.fuse_layers.2.3.0.weight: copying a param with shape torch.Size([64, 128, 1, 1]) from checkpoint, the shape in current model is torch.Size([128, 256, 1, 1]).
size mismatch for stage4.0.fuse_layers.2.3.1.weight: copying a param with shape torch.Size([64]) from checkpoint, the shape in current model is torch.Size([128]).
size mismatch for stage4.0.fuse_layers.2.3.1.bias: copying a param with shape torch.Size([64]) from checkpoint, the shape in current model is torch.Size([128]).
size mismatch for stage4.0.fuse_layers.2.3.1.running_mean: copying a param with shape torch.Size([64]) from checkpoint, the shape in current model is torch.Size([128]).
size mismatch for stage4.0.fuse_layers.2.3.1.running_var: copying a param with shape torch.Size([64]) from checkpoint, the shape in current model is torch.Size([128]).
size mismatch for stage4.0.fuse_layers.3.0.0.0.weight: copying a param with shape torch.Size([16, 16, 3, 3]) from checkpoint, the shape in current model is torch.Size([32, 32, 3, 3]).
size mismatch for stage4.0.fuse_layers.3.0.0.1.weight: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage4.0.fuse_layers.3.0.0.1.bias: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage4.0.fuse_layers.3.0.0.1.running_mean: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage4.0.fuse_layers.3.0.0.1.running_var: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage4.0.fuse_layers.3.0.1.0.weight: copying a param with shape torch.Size([16, 16, 3, 3]) from checkpoint, the shape in current model is torch.Size([32, 32, 3, 3]).
size mismatch for stage4.0.fuse_layers.3.0.1.1.weight: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage4.0.fuse_layers.3.0.1.1.bias: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage4.0.fuse_layers.3.0.1.1.running_mean: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage4.0.fuse_layers.3.0.1.1.running_var: copying a param with shape torch.Size([16]) from checkpoint, the shape in current model is torch.Size([32]).
size mismatch for stage4.0.fuse_layers.3.0.2.0.weight: copying a param with shape torch.Size([128, 16, 3, 3]) from checkpoint, the shape in current model is torch.Size([256, 32, 3, 3]).
size mismatch for stage4.0.fuse_layers.3.0.2.1.weight: copying a param with shape torch.Size([128]) from checkpoint, the shape in current model is torch.Size([256]).
size mismatch for stage4.0.fuse_layers.3.0.2.1.bias: copying a param with shape torch.Size([128]) from checkpoint, the shape in current model is torch.Size([256]).
size mismatch for stage4.0.fuse_layers.3.0.2.1.running_mean: copying a param with shape torch.Size([128]) from checkpoint, the shape in current model is torch.Size([256]).
size mismatch for stage4.0.fuse_layers.3.0.2.1.running_var: copying a param with shape torch.Size([128]) from checkpoint, the shape in current model is torch.Size([256]).
size mismatch for stage4.0.fuse_layers.3.1.0.0.weight: copying a param with shape torch.Size([32, 32, 3, 3]) from checkpoint, the shape in current model is torch.Size([64, 64, 3, 3]).
size mismatch for stage4.0.fuse_layers.3.1.0.1.weight: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for stage4.0.fuse_layers.3.1.0.1.bias: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for stage4.0.fuse_layers.3.1.0.1.running_mean: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for stage4.0.fuse_layers.3.1.0.1.running_var: copying a param with shape torch.Size([32]) from checkpoint, the shape in current model is torch.Size([64]).
size mismatch for stage4.0.fuse_layers.3.1.1.0.weight: copying a param with shape torch.Size([128, 32, 3, 3]) from checkpoint, the shape in current model is torch.Size([256, 64, 3, 3]).
size mismatch for stage4.0.fuse_layers.3.1.1.1.weight: copying a param with shape torch.Size([128]) from checkpoint, the shape in current model is torch.Size([256]).
size mismatch for stage4.0.fuse_layers.3.1.1.1.bias: copying a param with shape torch.Size([128]) from checkpoint, the shape in current model is torch.Size([256]).
size mismatch for stage4.0.fuse_layers.3.1.1.1.running_mean: copying a param with shape torch.Size([128]) from checkpoint, the shape in current model is torch.Size([256]).
size mismatch for stage4.0.fuse_layers.3.1.1.1.running_var: copying a param with shape torch.Size([128]) from checkpoint, the shape in current model is torch.Size([256]).
size mismatch for stage4.0.fuse_layers.3.2.0.0.weight: copying a param with shape torch.Size([128, 64, 3, 3]) from checkpoint, the shape in current model is torch.Size([256, 128, 3, 3]).
size mismatch for stage4.0.fuse_layers.3.2.0.1.weight: copying a param with shape torch.Size([128]) from checkpoint, the shape in current model is torch.Size([256]).
size mismatch for stage4.0.fuse_layers.3.2.0.1.bias: copying a param with shape torch.Size([128]) from checkpoint, the shape in current model is torch.Size([256]).
size mismatch for stage4.0.fuse_layers.3.2.0.1.running_mean: copying a param with shape torch.Size([128]) from checkpoint, the shape in current model is torch.Size([256]).
size mismatch for stage4.0.fuse_layers.3.2.0.1.running_var: copying a param with shape torch.Size([128]) from checkpoint, the shape in current model is torch.Size([256]).
Which is the HardNet architecture implemeted in this library?
Is it FC-HarDNet-70 from
https://github.com/PingoLH/FCHarDNet
RuntimeError: invalid argument 0: Sizes of tensors must match except in dimension 0. Got 366 and 335 in dimension 2 at /opt/conda/conda-bld/pytorch_1556653183467/work/aten/src/TH/generic/THTensor.cpp:711
in cityscapes datasets，a error happened . assert: DataLoader worker process 0 in line 57 assert (value in self._mapping). I changed it in line 51 to self._mapping = np.array(range(0, len(self._key))).astype('int32'). It works! Is it right?
python demo.py --config-file configs/pascal_voc_deeplabv3_plus.yaml --input_img '2012_004317.jpg'
Thanks for your excellent work!
state_dict_to_load = torch.load(cfg.TRAIN.PRETRAINED_MODEL_PATH)
File "/usr/local/lib/python3.6/site-packages/torch/serialization.py", line 382, in load
f = open(f, 'rb')
FileNotFoundError: [Errno 2] No such file or directory: 'runs/checkpoints/DeepLabV3_Plus_mobilenet_v2_coco_2019-12-03-15-40/best_model.pth
Thanks for the great work! I'm trying to use the multi-gpu training but got an error at line.
RuntimeError: Tensors must be CUDA and dense
How can i finetune this models??? @LikeLy-Journey
when use downloaded MobileNet v2,
got flops and params error: Given groups=1, weight of size 182 728 3 3, expected input[1, 96, 64, 128] to have 728 channels, but got 96 channels instead
and then in mobilenet.py line 134, in forward: x = self.conv1(x),
got UnboundLocalError: local variable 'total_ops' referenced before assignment
Other models such as use xception65 as backbone are doing fine
Hey,
I would like to know which backbones are compatible with this arch. Is this arch returning image of same scale as original??
Thanks for your excellent work!
Will you reproduce mobilenetv3+LR-ASPP ?
I have looked in your models and don´t know which one is 100 layers tiramisu
my test code like this,. The speed of deeplabv3-pluse-mobilenetv2(use ASPP and Decoder) is faster than the deeplabv3-pluse-mobilenetv2(doesn't use ASPP and Decoder). Do somebody know why?
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
Django
The Web framework for perfectionists with deadlines.
Laravel
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
Microsoft
Open source projects and samples from Microsoft.
Google
Google ❤️ Open Source for everyone.
Alibaba
Alibaba Open Source for everyone
D3
Data-Driven Documents codes.
Tencent
China tencent open source team.