train error

Hi.
I'm trying to train with your code and I'm getting an AssertionError.
The training was fine when I ran train.py first, but for some reason the training was aborted, when I ran train.py again and loaded the latest_ckpt.pth that was saved at the filrst training, I got an error.

commond: python tools/train.py -f exps/example/custom/yolox_m_table.py -d 1 -b 2 --fp16 -o --resume YOLOX_outputs/yolox_m_table/latest_ckpt.pth
error: Traceback (most recent call last): File "tools/train.py", line 87, in <module> exp.merge(args.opts) File "/home/Data/Projects/YOLOX-main/yolox/exp/base_exp.py", line 64, in merge assert len(cfg_list) % 2 == 0 AssertionError

and this is my config:

class Exp(MyExp):
    def __init__(self):
        super(Exp, self).__init__()
        # ---------------- model config ---------------- #
        self.num_classes = 2
        self.depth = 0.67
        self.width = 0.75
        self.exp_name = os.path.split(os.path.realpath(__file__))[1].split(".")[0]

        # ---------------- dataloader config ---------------- #
        self.data_num_workers = 2
        self.input_size = (960, 960)  # (height, width)
        self.data_dir = "/home/Data/Datasets/Table_Detection/coco"
        self.train_ann = "instances_train2017.json"
        self.val_ann = "instances_val2017.json"

        # --------------  training config --------------------- #
        self.max_epoch = 100
        self.basic_lr_per_img = 0.01 / 2.0
        self.no_aug_epochs = 15

        # -----------------  testing config ------------------ #
        self.test_size = (960, 960)

hope for your reply.

Start with tools/train.py and follow the resume command into yolox/exp/base_exp.py, especially exp.merge(args.opts), then reproduce the AssertionError using the command and config shown. Done means identifying why the resume invocation produces an invalid options list and documenting or correcting the supported command, with a regression check if the relevant test entry point is found.