FID without ground truth lengthes

[RohollahHS]

Hi Ekkasit,

Can you please share the code to calculate the FID score without ground truth lengthes?
I think the current code in GPT_eval_multi.py calculated the FID score with ground truth lengthes.

Thanks

[exitudio]

Hi Seyed,
I used pretrain legnth estimator from T2M (the proposed model from HumanML3D dataset paper). Please see "Train text2length model:" section.

Also, you can see from our unclean version . Please see: utils.eval_trans.py. And change is_pred_len=False to True.
The pretrained weights should be in length_est_bigru if you already set up HumanML3D.

[RohollahHS]

Thanks a lot