Hyuntae Park

nan loss while training

1

I use CUDA_VISIBLE_DEVICES=2 python tools/train_net.py --config-file "configs/e2e_mask_rcnn_R_101_FPN_1x.yaml" --skip-test SOLVER.IMS_PER_BATCH 2 SOLVER.BASE_LR 0.0025 SOLVER.MAX_ITER 720000 SOLVER.STEPS "(480000, 640000)" TEST.IMS_PER_BATCH 1 MODEL.RPN.FPN_POST_NMS_TOP_N_TRAIN 2000 this command to follow your instruction and I use...

About Novel, Unique score

1

Hello, While reproducing KM-BART, I realize I cannot get the same score as proposed in the paper. Why?? Is there another script used to compute novel, unique scores?

inaccurate VQA score

3

When I finetune the model with VQAv2, It produces a lower score than that proposed in the paper. (43%) Can you explain where this difference comes from and how to...

Hi! Following the experimental method shown in the SELFormer paper (Figure 4), we would like to analyze the drug discovery capabilities of various models. However, the ChemBL dataset and their...