[{"data":1,"prerenderedAt":259},["ShallowReactive",2],{"project-14855295":3},{"id":4,"name":5,"line":6,"line_source":7,"title":8,"url":9,"domain":10,"maker":11,"posted_at":12,"year":13,"points":14,"comments":15,"rank_in_year":16,"shown_in_year":17,"top":18,"tier":19,"category":20,"subcategory":21,"format":22,"keywords":23,"reception":28,"sentiment":29,"stance_mix":30,"scored_comments":35,"aspects":36,"strip":40,"post_text":43,"what":44,"said":63,"maker_in_thread":170,"shape":175,"grid":178,"attempts":216,"same_week":221,"maker_shows":176,"maker_hidden":251,"reader_version":252,"card":253},"14855295","Juggernaut","Experimental Neural Network in Rust","title","Show HN: Juggernaut – Experimental Neural Network in Rust","http:\u002F\u002Fjuggernaut.rs\u002F","juggernaut.rs","afshinmeh",1501068942,2017,54,19,755,2242,34,"notable","ai","neural network library","library",[24,25,26,27],"neural network","rust","machine learning","deep learning","roasted",-0.208,{"positive":31,"neutral":32,"mixed":33,"critical":34},1,2,0,4,7,{"idea":37,"design":38,"execution":39},{"positive":31,"critical":31},{"positive":33,"critical":31},{"positive":33,"critical":32},[31,33,41,33,33,33,33,33,33,31,33,42,41,31,33,33,33,31,33],3,5,null,{"problem":43,"audience":43,"pricing":43,"motivation":45,"time_to_build":43,"stack":55,"openness":61,"stage":62,"team_size":61},{"value":46,"source":47,"evidence":48},"Learning and gathering feedback on an early-stage project","maker",[49,52],{"id":50,"by":11,"text":51},"14856154","No not really no! I'm trying to gather some information and feedbacks about my project. I don't care (at this stage of the project) to be on the first page of HN or not, as long…",{"id":53,"by":11,"text":54},"14856146","thanks for your suggestions, shadowmint. I tried to mention it is \"experimental\", not a production ready NN. I'm working on adding more features to this project but for sure, it…",[56],{"value":57,"source":47,"evidence":58},"Rust",[59],{"id":60},"post","unknown","prototype",{"praise":64,"objections":73,"requests":103,"questions":120,"alternatives":140},[65],{"value":66,"source":67,"evidence":68},"Appreciation for projects that start small and can be followed","commenter",[69],{"id":70,"by":71,"text":72},"14858996","blahman2","Keep going! I really like being able to follow projects that start small, as opposed to 'here is my 10000 line toy project. That being said, you will probably get some flak,…",[74,84,89,96],{"value":75,"source":67,"evidence":76},"Too early and raw to be called experimental or shown on HN",[77,81],{"id":78,"by":79,"text":80},"14856089","shadowmint","There's nothing to see here. Trivial NN implementations are a dime a dozen, and this one is no different. It's just a partial work in progress; it's not a 'Show HN'; it's just a…",{"id":82,"by":79,"text":83},"14856193","Calling this an experimental neutral network is like me calling an opengl demo that renders a few flat shaded polygons an 'experimental 3d engine'. It's not. I'm sorry if that…",{"value":85,"source":67,"evidence":86},"Trivial implementation with no real capability, such as MNIST classification",[87,88],{"id":78,"by":79,"text":80},{"id":82,"by":79,"text":83},{"value":90,"source":67,"evidence":91},"Confusing API: training data inside network and activation fixed per network",[92],{"id":93,"by":94,"text":95},"14855836","yorwba","I'm confused by several of the API choices in the example. Why is the training set part of the network? I would have expected it to be a parameter to the train() function. Same…",{"value":97,"source":67,"evidence":98},"Rust offers little advantage over Python\u002FTensorFlow for neural networks",[99],{"id":100,"by":101,"text":102},"14855767","amelius","I'm a bit confused. How useful is this if: - Rust cannot compile to the GPU - Neural network programs are usually not large and therefore do not need the type safety that Rust…",[104,111,115],{"value":105,"source":67,"evidence":106},"Move training data to the train method",[107,108],{"id":93,"by":94,"text":95},{"id":109,"by":11,"text":110},"14856172","Thanks for your suggestions, yorwba. I do see what you mean about the API and I will work on this. For now, if you want to train a model, you need to pass data to the NN struct…",{"value":112,"source":67,"evidence":113},"Make activation a per-layer property",[114],{"id":93,"by":94,"text":95},{"value":116,"source":67,"evidence":117},"Classify MNIST before showing it",[118,119],{"id":78,"by":79,"text":80},{"id":82,"by":79,"text":83},[121,128],{"value":122,"source":67,"evidence":123},"What is the goal of the project?",[124],{"id":125,"by":126,"text":127},"14858779","bmh100","What is your goal in writing this project?",{"value":129,"source":67,"evidence":130},"Is Rust useful for neural networks given no GPU compilation?",[131,132,136],{"id":100,"by":101,"text":102},{"id":133,"by":134,"text":135},"14856512","pjmlp","> Rust cannot compile to the GPU Rust might not be able to compile to the GPU today, but any language can target the GPU, it is only a matter of tooling support.",{"id":137,"by":138,"text":139},"14857256","kersny","#1 is not strictly true, although it's not necessarily useful yet: https:\u002F\u002Fgithub.com\u002Fjaparic\u002Fnvptx#nvptx",[141,149,156,166],{"value":142,"source":67,"evidence":143},"TensorFlow and Keras",[144,145],{"id":100,"by":101,"text":102},{"id":146,"by":147,"text":148},"14858721","danieldk","which gives you quicker access to the new stuff, but the effort is about the same as writing a new implementation in another language Not really. You can just serialize a…",{"value":150,"source":67,"evidence":151},"PyTorch",[152],{"id":153,"by":154,"text":155},"14855807","mi_lk","> All cool neural network research is done on Keras\u002FTensorflow Add in PyTorch if you are really talking about nn research.",{"value":157,"source":67,"evidence":158},"Leaf",[159,162],{"id":160,"by":94,"text":161},"14855897","> - Neural network programs are usually not large and therefore do not need the type safety that Rust offers TensorFlow Fold [1] uses static type checking at model compilation…",{"id":163,"by":164,"text":165},"14856477","hobofan","Author of Leaf here. Yes, Leaf has been dead for more then a year. There recently has been an effort by some community members to pick up development of Leaf under a new name:…",{"value":167,"source":67,"evidence":168},"TensorFlow Fold",[169],{"id":160,"by":94,"text":161},{"comments":34,"answered":31,"key":171},[172,173,174],{"id":109,"text":110},{"id":50,"text":51},{"id":53,"text":54},{"commenters":176,"max_depth":34,"maker_replies":34,"first_comment_mins":177,"lifespan_hours":35},13,81,[179,181,183,185,187,188,190,193,194,195,197,199,201,203,205,209,211,212,213],[100,33,31,180,101,33],"project\u002Fcritical\u002Fidea",[153,33,33,182,154,33],"commenter\u002Fneutral\u002Fnone",[160,33,41,184,94,33],"project\u002Fneutral\u002Fidea",[163,33,33,186,164,33],"other\u002Fneutral\u002Fnone",[133,33,33,182,134,33],[189,33,33,182,164,33],"14856533",[191,33,33,182,192,33],"14858265","liuliu",[146,33,33,182,147,33],[137,33,33,186,138,33],[93,31,31,196,94,33],"project\u002Fcritical\u002Fdesign",[109,31,33,198,11,31],"maker\u002Fneutral\u002Fdesign",[70,32,42,200,71,33],"maker\u002Fpositive\u002Fidea",[125,41,41,202,126,33],"maker\u002Fneutral\u002Fidea",[78,34,31,204,79,33],"project\u002Fcritical\u002Fexecution",[206,34,33,207,208,33],"14856130","other\u002Fcritical\u002Fnone","overcast",[50,34,33,210,11,31],"maker\u002Fneutral\u002Fnone",[53,34,33,210,11,31],[82,34,31,204,79,33],[214,34,33,215,11,31],"14856208","maker\u002Fneutral\u002Fexecution",[217,220],{"id":218,"t":219,"points":41,"comments":33},"14837130",1500890492,{"id":4,"t":12,"points":14,"comments":15},[222,227,232,237,242,247],{"id":223,"name":224,"points":225,"category":226},"14821478","$200 Solar Self-Sufficiency  Without Your Landlord Noticing",491,"infrastructure",{"id":228,"name":229,"points":230,"category":231},"14859084","Chromeless",385,"dev_tools",{"id":233,"name":234,"points":235,"category":236},"14853212","Musikcube",340,"creative",{"id":238,"name":239,"points":240,"category":241},"14841859","Conference List",296,"productivity",{"id":243,"name":244,"points":245,"category":246},"14854915","Teachcraft",212,"languages",{"id":248,"name":249,"points":250,"category":231},"14821186","Mesh IDE",199,false,"v2",{"id":254,"name":5,"line":6,"ai":251,"cat":20,"year":13,"points":14,"comments":15,"rec":28,"s":255,"top":18,"kw":256,"maker":11,"domain":10,"tier":19,"attempts":32,"strip":257,"url":9,"mix":258,"praise":66,"objection":75,"sub":21},14855295,-0.21,[24,25,26,27],"1030000001053100010",[31,32,33,34],1794036557662]