SLDGroup / MERIT

41 stars 3 forks source link

question #8

Open liubaoning111 opened 4 months ago

liubaoning111 commented 4 months ago

Testing performance in val model: mean_dice : 0.726044, best_dice : 0.713541 save model to model_pth/MERIT_Cascaded_Small_loss_MUTATION_w3_7_Synapse256/MERIT_Cascaded_Small_loss_MUTATION_w3_7_pretrain_epo300_bs16_lr0.0001_256_s2222_run104504/best.pth 7%|██▏ | 22/300 [16:36<3:26:16, 44.52s/it]iteration 1000, epoch 22 : loss : 2.086078, lr: 0.000100

Thank you very much for your work! I found this problem when I ran the program again, meadn_dice>best_dice, may I ask what caused it? And more than once, meadn_dice>best_dice