MTtrans Model weight

<b>About this figshare deposit:</b>This figshare item is to store the weights of trained models. For source code and reproducibility about the study, please visit our github repository: https://github.com/holab-hku/MTtrans.<br><b>The naming of classes of model</b>: <b>M</b> stands for MPRA (Sample et al). <b>R</b> stands for ribosome profiling (Cao et al). As MTtrans is a multi-task model which aims to integrate different techniques, different combination of M and R leads to many different model classes. For example, <code><strong>3M</strong></code> means using 3 MPRA datasets to collectively train one MTtrans model, the resulting model will have 3 GRU tower. <code><strong>sMR</strong></code> means a sub-sampled MPRA was used together with a RP dataset.<b>Model checkpoints and configs</b>files name after <code>.pth</code> is the pytorch checkpoint file and <code>_cv%d</code> stands for different training repeats. Those name after <code>.ini</code> are the configuration files recording the hyper parameters and all other training settings.<code><strong>small_repective_filed_strides1113-model_best_cvx.pth</strong></code><b> </b>: the 3M model in the paper. The training process of model involves 3 MPRA datasets and <code><strong>V_PC3_RMSprop-model_best_cvx.pth</strong></code>: model sMR in the paper, which integrates a sub-sampled MRPA and the PC3 datasets<b>V_muscle_RMSprop-model_best_cvx.pth </b>: sMR model which use muscle RP dataset.<b>V_293_RMSprop-model_best_cvx.pth</b> : sMR model which trains on HEK293 RP dataset.