Berthy

> **Hello, I have the same problem. Have you solved it? This is my captions:** > > ``` > ID:205230 sentence: wipe hull rubs [ kids ] skating carved wipe...

A post in [this issue](https://github.com/longcw/yolo2-pytorch/issues/55#issuecomment-429763163) helped me. Before `running pooling/roi_align/make.sh`, add the following lines to `pooling/roi_align/make.sh` (right after line 10): ``` export PATH=/usr/local/cuda-8.0/bin${PATH:+:${PATH}} export CPATH=/usr/local/cuda-8.0/include${CPATH:+:${CPATH}} export LD_LIBRARY_PATH=/usr/local/cuda-8.0/lib64${LD_LIBRARY_PATH:+:${LD_LIBRARY_PATH}} ``` You might...

Thanks for the quick and detailed response!

For (2), I think the authors apply the normalization factor before taking the gradient. If you look at `ConditioningMethod.grad_and_value` ([here](https://github.com/DPS2022/diffusion-posterior-sampling/blob/82f313357accce87bd11c0f8d845ddb01ffee4e3/guided_diffusion/condition_methods.py#L31)), they take the gradient of the norm, not the norm...

@claroche-r thanks so much for clarifying that!