Skip to content

Latest commit

 

History

History
16 lines (12 loc) · 485 Bytes

eval.md

File metadata and controls

16 lines (12 loc) · 485 Bytes
Raw

Evaluation

Run the commands below to evaluate a pretrained model.

python eval.py --cfg configs/config.yaml

Change the TRAIN.PRETRAINED field of the config file to the checkpoint you would like to evaluate. You should be able to obtain the output below:

# TRAIN.PRETRAINED = 'data/vibe_data/vibe_model_wo_3dpw.pth.tar'
...Evaluating on 3DPW test set...
MPJPE: 93.5881, PA-MPJPE: 56.5608, PVE: 113.4118, ACCEL: 27.1242, ACCEL_ERR: 27.9877