load data

[SaidaSaad]

Hello

I would like to ask about the possibility to train using only the absent Kps. So can you please let me know how can I change in the function load_data_and_vocab(opt, load_train=True) in case of "one to many" to load only absent KPs ? Thanks

[kenchan0226]

You need to remove all the present keyphrases from the training data before running the preprocess.py. You need to remove the delimiter as well since you don't have present keyphrases.

[SaidaSaad]

Hello Kechan , Thanks you for your replies , I really appreciate. I have question about the reinforcement that may be not relate to . In case if I have another model that used to distinguish between the fake and true keypharse . This model I should use to evaluate the generate keypharses regarding this . I face problem in design the reward function . because this model gave me hidden states, logits and loss for each batch. I do not know which one I should use to calculate the rewards for each batch. I hope you can help me in that.