Recently, I have been using the pre-trained embeddings of promoter sequences as generation conditions based on your model for training.
Therefore, due to the slight differences in the models, I am currently unsure about the approximate performance of your model on the validation set in terms of val_recovery, val_loss, and val_perplexity.
Could you share some information with me? My results list below
Hi, Hannes
Recently, I have been using the pre-trained embeddings of promoter sequences as generation conditions based on your model for training.
Therefore, due to the slight differences in the models, I am currently unsure about the approximate performance of your model on the validation set in terms of val_recovery, val_loss, and val_perplexity.
Could you share some information with me? My results list below
