| python ../../finetune_classification.py \ | |
| --n_batch 32 \ | |
| --n_layer 24 \ | |
| --n_head 16 \ | |
| --n_embd 1024 \ | |
| --max_len 202 \ | |
| --d_dropout 0.1 \ | |
| --dropout 0.1 \ | |
| --lr_start 1e-7 \ | |
| --lr_multiplier 1 \ | |
| --max_epochs 500 \ | |
| --num_feats 32 \ | |
| --smi_ted_version 'v2' \ | |
| --model_path '../' \ | |
| --ckpt_filename 'smi-ted-Large_11.pt' \ | |
| --data_root '../../moleculenet/hiv' \ | |
| --dataset_name hiv \ | |
| --measure_name 'HIV_active' \ | |
| --checkpoints_folder './checkpoints_hiv_1e-7' \ | |
| --loss_fn 'crossentropy' \ | |
| --target_metric 'roc-auc' \ | |
| --n_output 2 \ | |
| --save_ckpt 1 \ | |
| --start_seed 0 \ | |
| --train_decoder 0 \ |