search

Mohamedelrefaie / DrivAerNet

A Large-Scale Multimodal Car Dataset with Computational Fluid Dynamics Simulations and Deep Learning Benchmarks
Other
139 stars 20 forks source link

Know details about trainning #18

Open LilaKen opened 2 weeks ago

LilaKen commented 2 weeks ago

when i trainning the diveraernet in RegDGCNN, the experimental R2 score is never higher than zero, whatever many attempts about parameter，chould u expalin this？

LilaKen commented 1 week ago

i found drivaernet’s R2 score is different R2 score， and we find that other R2 calculation methods are different from those in the original paper. u also can use pred and target to calculate your R2 score.

JJ-interesting commented 6 days ago

i found drivaernet’s R2 score is different R2 score， and we find that other R2 calculation methods are different from those in the original paper. u also can use pred and target to calculate your R2 score.

Thank you for your response. Could you explain in detail how the R² score is calculated and what the differences are? In the code, it seems like the total R² score is calculated and then the average R² score is computed. It appears to be no different from the traditional R² score.

JJ-interesting commented 3 days ago

Did you manage to reproduce the results from the paper? If so, could you share them? These are the results I obtained from my training. The R² score has turned positive but hasn't reached the 0.901 reported in the paper. Total training time: 11301.56s Model saved to models/cdPrediction DrivAerNet r2 100epochs 5k final model.pth Testing the final model: Test MSE: 0.000292, TeSt MAE: 0.013641, MaX MAE: 0.022632, Test R2: 0.5032 Total inference time: 1.21s for 592 samples Testing the best model: Test MSE: 0.000229, TeSt MAE: 0.011788, MaX MAE: 0.022385, Test R2: 0,6315 Total inference time: 1,10s for 592 samples