How to further train on our own dataset?

[pranavkamath14]

Hello, Thank you so much for this amazing github repo. I wanted to use the pre-trained model and further train on my own dataset of videos and corresponding captions that I have. Do you have any suggestions on how I can do this?

[qt2139]

same question

[ttengwang]

1) prepare the video features (.npy files) following TSP and convert_tsp_h5_to_npy.py. 2) construct the follow annoations of your dataset. train_caption_file, val_caption_file, visual_feature_folder, gt_file_for_eval, gt_file_for_para_eval, dict_file 3) To further finetune the pre-trained PDVC (TSP), run python train.py --cfg_path=anet_tsp_pdvc.yml --pretrain=full --pretrain_path=path/to/ckpt