Prediction values changes every run

[NasreddineD]

Hi, with DeepLearning models (even for deterministic model), each time I run the prediction (even without retraining), the results are always different. The MAPE calculation for each time I run the prediction can go from 3% to 20% which is inexploitable... Is it normal? Am I doing something wrong?

Thank you very much.

[hrzn]

Hi, could you please post a minimal self-contained snippet showing the case where the same trained deterministic model produces different predictions? Thanks.