Process finished with exit code 139 (interrupted by signal 11: SIGSEGV)

[as754770178]

I exec train_net.py, the parameter config-file is maskrcnn-benchmark/configs/e2e_mask_rcnn_R_50_FPN_1x.yaml. I change maskrcnn-benchmark/configs/e2e_mask_rcnn_R_50_FPN_1x.yaml as below:

DATASETS:
  TRAIN: ("coco_2014_train",)
  TEST: ("coco_2014_val",)

change config.py as below:

       "coco_2014_train": {
            "img_dir": "/home/xtz/hard_disk/disk_2/coco/images/train2014",
            "ann_file": "/home/xtz/hard_disk/disk_2/coco/annotations/instances_train2014.json"
        },
        "coco_2014_val": {
            "img_dir": "/home/xtz/hard_disk/disk_2/coco/images/val2014",
            "ann_file": "/home/xtz/hard_disk/disk_2/coco/annotations/instances_val2014.json"
        },

I fixed IndexError: list index out of range error refer to #725.

when compute loss loss_dict = model(images, targets) in trainer.py, the program report error:

Process finished with exit code 139 (interrupted by signal 11: SIGSEGV)

[Darshan2701]

@as754770178 This happens if the learning rate is too high i guess, please lower the base learning rate in case if you are training the model from scratch.