[{"data":1,"prerenderedAt":303},["ShallowReactive",2],{"project-15076727":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":42,"post_text":44,"what":45,"said":64,"maker_in_thread":154,"shape":164,"grid":168,"attempts":266,"same_week":267,"maker_shows":41,"maker_hidden":295,"reader_version":296,"card":297},"15076727","got-book-6","RNN that generates Game of Thrones text","title","Show HN: RNN that generates Game of Thrones text","https:\u002F\u002Fwww.github.com\u002Fzackthoutt\u002Fgot-book-6","github.com","zthoutt",1503439108,2017,143,38,274,2242,13,"notable","ai","text generation","other",[24,21,25,26,27],"rnn","lstm","game of thrones","markov chain","mixed",-0.024,{"positive":31,"neutral":32,"mixed":33,"critical":34},5,9,0,7,21,{"idea":37,"performance":39,"execution":40},{"positive":38,"critical":38},2,{"positive":33,"critical":33},{"positive":33,"critical":41},1,[41,33,33,33,33,33,33,41,33,43,41,43,43,41,33,31,33,43,33,31,41,41,31,33,33,33,33,41,33,33,43,43,31,33,31,43,43,43],3,null,{"problem":44,"audience":44,"pricing":44,"motivation":44,"time_to_build":44,"stack":46,"openness":62,"stage":63,"team_size":63},[47,54],{"value":48,"source":49,"evidence":50},"TensorFlow 1.0.0","maker",[51],{"id":52,"by":11,"text":53},"15077095","Yep, I was using TF 1.0.0. I will probably upgrade in the future and will keep this in mind.",{"value":55,"source":56,"evidence":57},"LSTM recurrent neural network","commenter",[58],{"id":59,"by":60,"text":61},"15077057","minimaxir","It should be noted that per the included Jupyter Notebook, the resulting network took 24 hours to train, and the resulting network (of a 3-layer 512 cell LSTM) would be hundreds…","open_source","unknown",{"praise":65,"objections":89,"requests":116,"questions":135,"alternatives":144},[66,81],{"value":67,"source":56,"evidence":68},"Funny and surprisingly coherent output",[69,73,77],{"id":70,"by":71,"text":72},"15076848","notamy","This is amazing. I wonder if it'd be possible to actually get a coherent plot out of it while retaining the humorous aspects of it.",{"id":74,"by":75,"text":76},"15076865","SwellJoe","The RNN seems to agree with the most popular fan theory about what will happen with Jaime and Cersei. So, it's not entirely crazy. Nonetheless, I don't think GRR Martin has to…",{"id":78,"by":79,"text":80},"15076910","jaz46","This is hilarious! We did a similar project last year using the scripts from the show. http:\u002F\u002Fdocs.pachyderm.io\u002Fen\u002Flatest\u002Fexamples\u002Ftensor_flow\u002Freadme.html Ours was not nearly as…",{"value":82,"source":56,"evidence":83},"Fun project to see",[84,85],{"id":74,"by":75,"text":76},{"id":86,"by":87,"text":88},"15077629","wbertelsen","Nice! I did a similar thing with a character-based rnn a few years ago, but its output was less coherent. I liked when it started spitting out character descriptions though:…",[90,105],{"value":91,"source":56,"evidence":92},"Copyright concerns over distributing the full book text in the data folder",[93,97,101],{"id":94,"by":95,"text":96},"15077701","DoctorNick","You need to take down the data folder now and rebuild the repository. You're going to get hit with a DMCA notice. It's not a matter of IF, it's a matter of WHEN. Publishers of…",{"id":98,"by":99,"text":100},"15076908","rlglwx","That data folder seems ripe for a DMCA takedown request.",{"id":102,"by":103,"text":104},"15078567","mrhigat4","You really should not upload other people's complete works while under copyright. Edit: Nice License > Copyright (c) 2017 Zack Thoutt",{"value":106,"source":56,"evidence":107},"Output is incoherent and looks like a Markov chain",[108,112],{"id":109,"by":110,"text":111},"15077429","Semiapies","It does seem like a lot of work to get output that looks like something from a simple Markov chain.",{"id":113,"by":114,"text":115},"15081208","donquichotte","With Markov Chains: \"Raved on fire, he clambered up her pallet in a swatch of blood and crept toward him shudder. She was too well Ghost back behind their crews could not lack…",[117,124],{"value":118,"source":56,"evidence":119},"Add an attention mechanism for better results",[120],{"id":121,"by":122,"text":123},"15079743","JustFinishedBSG","You definitely need to add an attention mechanism if you want better-than-markov-chain results. Found a project ;)",{"value":125,"source":56,"evidence":126},"Add more training data such as fan fiction or scripts",[127,131],{"id":128,"by":129,"text":130},"15077675","itschekkers","I wonder if there are any forums\u002Ffan fiction websites where you could find big chunks of prose to add to the training data. Might help bring the loss down a bit if you mix some in.",{"id":132,"by":133,"text":134},"15079853","alexcnwy","This brings up the interesting question of whether it's ok to publish the trained model and not the data? I'd imagine it's sufficiently derivative as to be allowed? What if you…",[136],{"value":137,"source":56,"evidence":138},"Is it legal to publish a trained model without the training data?",[139,140],{"id":132,"by":133,"text":134},{"id":141,"by":142,"text":143},"15083577","ebp234","I believe it would still be a derivative work. https:\u002F\u002Fwww.copyright.gov\u002Fcircs\u002Fcirc14.pdf",[145,149],{"value":146,"source":56,"evidence":147},"textgenrnn",[148],{"id":59,"by":60,"text":61},{"value":150,"source":56,"evidence":151},"Markov chain text generators",[152,153],{"id":109,"by":110,"text":111},{"id":113,"by":114,"text":115},{"comments":155,"answered":33,"key":156},4,[157,160,161],{"id":158,"text":159},"15077105","Haha, that's awesome! I was thinking about trying to combine the scripts and book text so that I can train on more data.",{"id":52,"text":53},{"id":162,"text":163},"15099012","A message to all of the vigilantes on HN: I'm sorry that the GOT text ended up in the repo. I actually worked on this project about a month ago and after showing it to a few…",{"commenters":165,"max_depth":34,"maker_replies":155,"first_comment_mins":166,"lifespan_hours":167},30,12,65,[169,171,175,178,181,183,185,187,190,193,195,197,201,203,205,208,210,212,216,217,218,219,221,225,229,232,236,238,240,244,247,248,250,251,254,256,260,264],[94,33,41,170,95,33],"maker\u002Fcritical\u002Fnone",[172,33,33,173,174,33],"15078221","commenter\u002Fneutral\u002Fnone","marssaxman",[176,33,33,173,177,33],"15078239","kbenson",[179,33,33,173,180,33],"15081195","EpicEng",[182,33,33,173,174,33],"15083346",[184,33,33,173,180,33],"15084364",[186,33,33,173,174,33],"15084474",[188,33,41,189,180,33],"15084524","project\u002Fcritical\u002Fnone",[191,33,33,173,192,33],"15085353","afarrell",[132,33,43,194,133,33],"project\u002Fneutral\u002Fidea",[141,33,41,196,142,33],"project\u002Fcritical\u002Fidea",[198,33,43,199,200,33],"15081055","project\u002Fneutral\u002Fnone","spapas82",[59,41,43,202,60,33],"project\u002Fneutral\u002Fperformance",[109,41,41,204,110,33],"project\u002Fcritical\u002Fexecution",[206,41,33,173,207,33],"15077647","visarga",[78,38,31,209,79,33],"project\u002Fpositive\u002Fnone",[158,38,33,211,11,41],"maker\u002Fpositive\u002Fidea",[213,43,43,214,215,33],"15076809","maker\u002Fneutral\u002Fexecution","yablak",[52,43,33,214,11,41],[74,155,31,209,75,33],[98,31,41,189,99,33],[102,220,41,189,103,33],6,[222,34,31,223,224,33],"15077228","project\u002Fpositive\u002Fidea","dfee",[226,34,33,227,228,33],"15077639","other\u002Fneutral\u002Fnone","thegeomaster",[230,34,33,227,231,33],"15078111","xirdstl",[233,234,33,227,235,33],"15078034",8,"ryanlol",[162,32,33,237,11,41],"maker\u002Fneutral\u002Fnone",[113,239,41,196,114,33],10,[241,242,33,227,243,33],"15077394",11,"yakult",[245,242,33,227,246,33],"15077515","gpm",[128,166,43,194,129,33],[249,166,43,199,246,33],"15077968",[70,18,31,223,71,33],[252,18,33,253,11,41],"15077076","maker\u002Fneutral\u002Fperformance",[86,255,31,209,87,33],14,[257,258,43,194,259,33],"15079293",15,"xiphias",[261,262,43,199,263,33],"15079415",16,"adtac",[121,265,43,202,122,33],17,[],[268,273,277,282,287,291],{"id":269,"name":270,"points":271,"category":272},"15074526","Champion Traveler",395,"consumer_social",{"id":274,"name":275,"points":276,"category":22},"15041620","Product Graveyard",265,{"id":278,"name":279,"points":280,"category":281},"15070686","Pdf-Bot",235,"dev_tools",{"id":283,"name":284,"points":285,"category":286},"15073519","YC Remix",132,"productivity",{"id":288,"name":289,"points":290,"category":281},"15043432","basic-build",100,{"id":292,"name":293,"points":294,"category":281},"15035747","Appster",79,false,"v2",{"id":298,"name":5,"line":6,"ai":295,"cat":20,"year":13,"points":14,"comments":15,"rec":28,"s":299,"top":18,"kw":300,"maker":11,"domain":10,"tier":19,"attempts":33,"strip":301,"url":9,"mix":302,"praise":67,"objection":91,"sub":21},15076727,-0.02,[24,21,25,26,27],"10000001031331050305115000010033505333",[31,32,33,34],1793990963144]