Xin Kong
Hi! I have a quick question. Is functorch able to train dynamic ensemble models in a batch training way? For example, I have several models of the same architecture with...
Hi! The same issue. Has anyone implemented this?
@vinayakkailas Hi I met the same problem. `ImportError: No module named 'fast_rcnn'` Did you solve this problem? Thank you!
Hi, for the pose info in JSON file is directly obtained from KITTI dataset ground truth data.
@anju000 Hi! I want to see the tracked points as well. Have you solved this problem? Thanks.
I make `pcl::PointCloud::Ptr previousModel_` be public and publish these points to the rviz with ROS. I notice that these points are exactly the points I add (using `addPoints`). When I...
./TIM-x86_64.AppImage /tmp/.mount_TIM-x8yX46iv/AppRun: 行 23: /tmp/.mount_TIM-x8yX46iv/usr/bin/wine: 没有那个文件或目录 Cleaning up @askme765cs 2018.7.17使用还是报错
@bkoyuncu Hi, here is what I did. 1. Init batch of models: models is a list of models with exact same structure. `fmodel, params, buffers = combine_state_for_ensemble(models)` `[p.requires_grad_() for p...
A link to my project [vMAP](https://github.com/kxhit/vMAP) where I used functorch vmap function for vectorised training of multiple tiny MLPs (NeRF). Thanks a lot to @zou3519 the amazing developer team of...
@amyeroberts @junwang-wish Hi I have the same issue with transformer==4.30.2. I found the preprocessing makes the difference. I tried 3 different ways to do the preprocessing and only the 3....