Hello, I am having difficulty creating language_model_learner and classifier
language_model_learner
- Suppose I create a LM using the signature below
learner = language_model_learner(data_lm, AWD_LSTM, pretrained=False,
pretrained_fnames=[‘name1’,‘name2’])
Since I am not able to download from internet I am specifying pretrained=False, but specifying the name in pretrained_fnames will load the pretrained LM and vocab ?
- Similarly for classifier after training LM
classifier = text_classifier_learner(data_lm, AWD_LSTM, pretrained=False) ,
how do you load the pre_trained model if you specify pretrained=False
I have to do this since I am behind firewall