search

airsplay / lxmert

PyTorch code for EMNLP 2019 paper "LXMERT: Learning Cross-Modality Encoder Representations from Transformers".
MIT License
933 stars 158 forks source link

cudaSuccess (9 vs. 0) #42

Closed volquelme closed 4 years ago

volquelme commented 4 years ago

hi, thanks for your sharing i was trying to extract object features on other dataset(images) but, i met this error i just updated my pytorch to 1.4 but still same error come out... how can i solve this? capture

volquelme commented 4 years ago

it might be caused by some images with too small resolution