• 1 Post
  • 109 Comments
Joined 6 years ago
cake
Cake day: July 17th, 2018

help-circle


  • @something_random_tho The pathfinding algorithm is in the client app (like Organic Maps) and the data are in openstreetmap. Different apps can find different paths and there is not much we can do about it, because changing the pathfinding algorithm could break it for many other places. That’s why I would rather modify the data, such as marking roads unsuitable for bikes, or making sure the nodes are properly connected, adding speed limits, road surface types… nudge the algo into the right path














  • @vrighter @ylai
    That is a really bad analogy. If the “compilation” takes 6 months on a farm of 1000 GPUs and the results are random, then the dataset is basically worthless compared to the model. Datasets are easily available, always were, but if someone invests the effort in the training, then they don’t want to let others use the model as open-source. Which is why we want open-source models. But not “openwashed” where they call it “open” for non-commercial, no modifications, no redistribution