search

facebookresearch / CutLER

Code release for "Cut and Learn for Unsupervised Object Detection and Instance Segmentation" and "VideoCutLER: Surprisingly Simple Unsupervised Video Instance Segmentation"
Other
918 stars 90 forks source link

Self-training error #13

Closed Greybeast closed 1 year ago

Greybeast commented 1 year ago

After having generated the json for the first round of training, I get this error. The given snippet asks to provide the train-dataset value for round1 training, but it seems script does not take train-dataset input. Is it the "test-dataset" that should be provided?

python train_net.py --num-gpus 8 \ --config-file model_zoo/configs/CutLER-ImageNet/cascade_mask_rcnn_R_50_FPN_self_train.yaml \ --train-dataset imagenet_train_r1 \ MODEL.WEIGHTS output/model_final.pth \ # load previous stage/round checkpoints OUTPUT_DIR output/self-train-r1/ # path to save checkpoints

Error: usage: train_net.py [-h] [--config-file FILE] [--resume] [--eval-only] [--num-gpus NUM_GPUS] [--num-machines NUM_MACHINES] [--machine-rank MACHINE_RANK] [--test-dataset TEST_DATASET] [--no-segm] [--dist-url DIST_URL] ... train_net.py: error: unrecognized arguments: --train-dataset

I have used a very very small subset of imagenet and have successfully executed till generation of cutler_imagenet1k_train_r1.json. Please help!

frank-xwang commented 1 year ago

Hi @Greybeast, we have updated the codes to fix this issue.

Also, you can remove the "--train-dataset imagenet_train_r1" command and modify the arguments in "model_zoo/configs/CutLER-ImageNet/cascade_mask_rcnn_R_50_FPN_self_train.yaml" directly. This is Detectron2's recommended way to change configurations.

Both methods should fix your bugs. Please let us know if they work. Thank you! XD

frank-xwang commented 1 year ago

Closing it now. Please feel free to reopen it if have further questions.