IOP

TAT: Preprocessing a dataset

python -m TAT.preprocessing --dataset CollegeMsg

TAT: Run the baseline model

python -m TAT.main --dataset CollegeMsg --model TAT --gpu 0

TAT: Run the sequence prediction model

python -m TAT.main_seq_pred --dataset CollegeMsg --model TAT --gpu 0

TAT: Run the prediction at timestep model

python -m TAT.main_pred_t --dataset CollegeMsg --model TAT --gpu 0

JODIE: Generate embeddings

python jodie.py --network CollegeMsg --model jodie --epochs 50

JODIE: IOP task (To use projected embeddings replace dyn_emb.py with dyn_emb_projected.py)

python dyn_emb.py --network CollegeMsg --model jodie --epochs 50

See individual READMEs for more details.