dimenwarper / molxspec Goto Github PK

When I run training_ setup. py file, the system prompts checkpoint: optional [any] = none Attributeerror: module 'Torch' has no attribute 'model'.
Have you encountered this problem before? If yes, do you know how to solve it?

These are the adduct types being considered by the models:

ADDUCTS = ['[M+H]+', '[M+Na]+', 'M+H', 'M-H', '[M-H2O+H]+', '[M-H]-', '[M+NH4]+', 'M+NH4', 'M+Na']

As you can see, there are repeated adducts when you remove the brackets, due to different notations. We should merge these adduct types as the bracket version have lower samples in the GNPS data and may be artifactually biasing the results. This of course means retraining the models.