about json file， the input size and crop size should based on what

[ShiMinghao0208]

my dataset image size is 256*256，and i dont know how to modifiy the json file

{
    "dataset": "BDCI",
    "method": "fcn_segmentor",
    "data": {
      "image_tool": "cv2",
      "input_mode": "BGR",
      "num_classes": 7,
      "label_list": [0, 1, 2, 3, 4, 5, 6, 255],
      "data_dir": "~/DataSet/BDCI",
      "workers": 8
    },
   "train": {
      "batch_size": 16,
      "data_transformer": {
        "size_mode": "fix_size",
        "input_size": [256, 256],
        "align_method": "only_pad",
        "pad_mode": "random"
      }
    },
    "val": {
      "batch_size": 4,
      "mode": "ss_test",
      "data_transformer": {
        "size_mode": "fix_size",
        "input_size": [256, 256],
        "align_method": "only_pad"
      }
    },
    "test": {
      "batch_size": 4,
      "mode": "ss_test",
      "out_dir": "~/DataSet/BDCI/seg_result/BDCI",
      "data_transformer": {
        "size_mode": "fix_size",
        "input_size": [256, 256],
        "align_method": "only_pad"
      }
    },
    "train_trans": {
      "trans_seq": ["random_resize", "random_crop", "random_hflip", "random_brightness"],
      "random_brightness": {
        "ratio": 1.0,
        "shift_value": 10
      },
      "random_hflip": {
        "ratio": 0.5,
        "swap_pair": []
      },
      "random_resize": {
        "ratio": 1.0,
        "method": "random",
        "scale_range": [0.5, 2.0],
        "aspect_range": [0.9, 1.1]
      },
      "random_crop":{
        "ratio": 1.0,
        "crop_size": [256, 256],
        "method": "random",
        "allow_outside_center": false
      }
    },
    "val_trans": {
      "trans_seq": []
    },
    "normalize": {
      "div_value": 255.0,
      "mean_value": [0.485, 0.456, 0.406],
      "mean": [0.485, 0.456, 0.406],
      "std": [0.229, 0.224, 0.225]
    },
    "checkpoints": {
      "checkpoints_name": "fs_baseocnet_BDCI_seg",
      "checkpoints_dir": "./checkpoints/BDCI",
      "save_iters": 500
    },
    "network":{
      "backbone": "deepbase_resnet101_dilated8",
      "multi_grid": [1, 1, 1],
      "model_name": "base_ocnet",
      "bn_type": "inplace_abn",
      "stride": 8,
      "factors": [[8, 8]],
      "loss_weights": {
        "corr_loss": 0.01,
        "aux_loss": 0.4,
        "seg_loss": 1.0
      }
    },
    "logging": {
      "logfile_level": "info",
      "stdout_level": "info",
      "log_file": "./log/BDCI/fs_baseocnet_BDCI_seg.log",
      "log_format": "%(asctime)s %(levelname)-7s %(message)s",
      "rewrite": true
    },
    "lr": {
      "base_lr": 0.01,
      "metric": "iters",
      "lr_policy": "lambda_poly",
      "step": {
        "gamma": 0.5,
        "step_size": 100
      }
    },
    "solver": {
      "display_iter": 10,
      "test_interval": 1000,
      "max_iters": 40000
    },
    "optim": {
      "optim_method": "sgd",
      "adam": {
        "betas": [0.9, 0.999],
        "eps": 1e-08,
        "weight_decay": 0.0001
      },
      "sgd": {
        "weight_decay": 0.0005,
        "momentum": 0.9,
        "nesterov": false
      }
    },
    "loss": {
      "loss_type": "fs_auxce_loss",
      "params": {
        "ce_weight": [0.8373, 0.9180, 0.8660, 1.0345, 1.0166, 0.9969, 0.9754,
                      1.0489, 0.8786, 1.0023, 0.9539, 0.9843, 1.1116, 0.9037,
                      1.0865, 1.0955, 1.0865, 1.1529, 1.0507],
        "ce_reduction": "elementwise_mean",
        "ce_ignore_index": -1,
        "ohem_minkeep": 100000,
        "ohem_thresh": 0.9
      }
    }
}

here is my json file, and when i try to train my dataset, there is such sizemisbatch error...like:


and so on,
environment should be satisfied：

this is my val error:

and the config.profile:

this is my log file screenshot:

[hsfzxjy]

@ShiMinghao0208 Hi. Seems that you are using HRNet-W48 as backbone model, but trying to load pretrained weights of HRNet-W32.

[ShiMinghao0208]

@ShiMinghao0208 Hi. Seems that you are using HRNet-W48 as backbone model, but trying to load pretrained weights of HRNet-W32.

Oh，I think one model can use several different pre-training models. I change the HRNet-W48 pretrained model, it does work. Very thank u.

[hsfzxjy]

They have different structure and thus the weights could not be shared.

[ShiMinghao0208]

They have different structure and thus the weights could not be shared.

ok，thanks. I will try other model，too. I am very sorry to bother u.

[ShiMinghao0208]

They have different structure and thus the weights could not be shared.

I want to try to use script val 1 after my training process，but I get a error about file dir error. The training process is normal,and I check my DATA_ROOT, I dont know why I get this error

[hsfzxjy]

They have different structure and thus the weights could not be shared.

I want to try to use script val 1 after my training process，but I get a error about file dir error. The training process is normal,and I check my DATA_ROOT, I dont know why I get this error

Could you share the training log? So that we could help figure out the problem.

[ShiMinghao0208]

They have different structure and thus the weights could not be shared.

I want to try to use script val 1 after my training process，but I get a error about file dir error. The training process is normal,and I check my DATA_ROOT, I dont know why I get this error

Could you share the training log? So that we could help figure out the problem.

They have different structure and thus the weights could not be shared.

I want to try to use script val 1 after my training process，but I get a error about file dir error. The training process is normal,and I check my DATA_ROOT, I dont know why I get this error

Could you share the training log? So that we could help figure out the problem.

OK, thank u very much. I upload the screenshot about my log and question in the issue.Does that help solve the problem?

[hsfzxjy]

Could you provide the script file (the .sh file) you are using? You may upload the file by attaching it to a comment. I guess there should be some mistakes in the validation command.

[ShiMinghao0208]

json_log_sh.zip

[ShiMinghao0208]

Could you provide the script file (the .sh file) you are using? You may upload the file by attaching it to a comment. I guess there should be some mistakes in the validation command.

OK， I upload the zip of my file(sh, json and so on).

[hsfzxjy]

Hi. You should add extra arguments --data_dir ${DATA_DIR} to the command, i.e., change it into:

${PYTHON} -u main.py --configs ${CONFIGS} --drop_last y \
                       --backbone ${BACKBONE} --model_name ${MODEL_NAME} --checkpoints_name ${CHECKPOINTS_NAME} \
                       --phase test --gpu 0 1 2 --resume ./checkpoints/BDCI/${CHECKPOINTS_NAME}_latest.pth \
                       --loss_type ${LOSS_TYPE} --test_dir ${DATA_DIR}/val/image \
                       --out_dir ${SAVE_DIR}${CHECKPOINTS_NAME}_val \
                       --data_dir ${DATA_DIR}

[ShiMinghao0208]

Hi. You should add extra arguments --data_dir ${DATA_DIR} to the command, i.e., change it into:

${PYTHON} -u main.py --configs ${CONFIGS} --drop_last y \
                       --backbone ${BACKBONE} --model_name ${MODEL_NAME} --checkpoints_name ${CHECKPOINTS_NAME} \
                       --phase test --gpu 0 1 2 --resume ./checkpoints/BDCI/${CHECKPOINTS_NAME}_latest.pth \
                       --loss_type ${LOSS_TYPE} --test_dir ${DATA_DIR}/val/image \
                       --out_dir ${SAVE_DIR}${CHECKPOINTS_NAME}_val \
                       --data_dir ${DATA_DIR}

OK，thanks very much. I will try it