quhezheng

I run into the same issue and exactly around 25000 steps. Both accelerator and brake are used at the maximum extent. The car stuck on the ground.

I think it should be 1. Don't know why 3

@AmirCognitive You are right, I run into same issue. OU output are very confusing, the steering come out mostly as positive value, the policy seldom try negtive value. The init...

I think you are right. Becuase reward is a scaler, which can not server a vector Q value

@evjlobanova Is there any reference document tell us how to interface with C++ or any other language?

Is there anyway to optimize the code to view large las file? It would be a pity if we can't use it on city scope point cloud navigation

@ppwwyyxx So nice to see you here, Tensorpack author, why this repo's performance differ so much from your samples in tensorpack. I don't see major difference but this repo collect...

@ppwwyyxx I failed to describe the issue clearly. The issue I run into this reop is the training best rewards is 30, no where to compare with the sample in...

@ICTwangbiao (classes+num_object*5) * side *side comes from the paper. But this code use 7 instead of 30 to represent one cell. [https://github.com/yeahkun/caffe-yolo/blob/master/src/caffe/layers/box_data_layer.cpp#L142](url) make it very clear.: > CHECK_EQ(count, locations *...