search

ProteomeTeam / NovoB

Bidirectional De Novo Peptide Sequencing Using a Transformer Model
2 stars 0 forks source link

What is the difference between with and without using --load_weights when using Prediction.py? #1

Closed irleader closed 4 months ago

irleader commented 4 months ago

There are two examples using Prediction.py"

python Prediction.py -m NovoBInit/ --load_weights TrainedModel/usingCasanovoDataSet/yeast/variables/variables python Prediction.py -m TrainedModel/usingCasanovoDataSet/yeast/

one with NovoBInit and --load weigths as well as the variables. The other with the whole folder of trained model.

Will the output be different or the same? if different, what is the difference. Thanks a lot.

pardess commented 4 months ago

NovoB is loaded model or model+weight. If you want to load the learned model, don't use the --load_weights option. Depending on your tensorflow version, if --save_model is not available, you must use --save_weights to save the weights. In this case, you can load the saved weights using --load_weights.