René's URL Explorer Experiment

{"@context":"https://schema.org","@type":"DiscussionForumPosting","headline":"RNN Models","articleBody":"Once we have an implementation of the Layer Class https://github.com/arrayfire/arrayfire_ml/issues/17 , the Optimizer class and the DataSet class we can go about creating RNN flavors. There are 3 models that should be implemented:\n- [ ] Vanilla RNN\n- [ ] LSTM\n- [ ] GRU\n\nThese will require the implementation of their derivatives and their forward prop values. \nCertain details to consider:\n- RNN's have a stack of weight matrices and bias' (not just 1 per Layer, thus the Layer needs to be general enough to handle this)\n- The optimization needs to be handled via two methods: \n  - [ ]  RTRL (real time recurrent learning) \u0026\n  - [ ]  BPTT (backprop through time)\n\nTo enable the above two methods of learning we should consider inheriting from Layer and implementing  a Recurrent Layer.\n","author":{"url":"https://github.com/jramapuram","@type":"Person","name":"jramapuram"},"datePublished":"2015-06-18T09:27:51.000Z","interactionStatistic":{"@type":"InteractionCounter","interactionType":"https://schema.org/CommentAction","userInteractionCount":12},"url":"https://github.com/20/arrayfire-ml/issues/20"}