search

airsplay / lxmert

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

Could you also provide the fine-tuned model? #40

Closed Princesston closed 4 years ago

Princesston commented 4 years ago

Thanks for the work, could you also provide the fine-tuned model for VQA, GQA and NLVR2? That would be very helpful, thanks.

airsplay commented 4 years ago

VQA: https://github.com/airsplay/lxmert/issues/7 GQA: https://github.com/airsplay/lxmert/issues/16

My disk got cleaned recently and thus do not have NLVR2 in hands...