[{"data":1,"prerenderedAt":125},["ShallowReactive",2],{"project-14947671":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":15,"aspects":32,"strip":34,"post_text":29,"what":36,"said":45,"maker_in_thread":67,"shape":69,"grid":72,"attempts":75,"same_week":86,"maker_shows":116,"maker_hidden":117,"reader_version":118,"card":119},"14947671","textgenrnn","Python module to easily generate text using a pretrained char-rnn","title","Show HN: Python module to easily generate text using a pretrained char-rnn","https:\u002F\u002Fgithub.com\u002Fminimaxir\u002Ftextgenrnn","github.com","minimaxir",1502116095,2017,14,1,1466,2242,66,"shown","ai","text generation","library",[24,21,25,26,27],"char-rnn","python","recurrent neural network","pretrained model","quiet",null,{"positive":15,"neutral":31,"mixed":31,"critical":31},0,{"idea":33},{"positive":15,"critical":31},[35],5,{"problem":29,"audience":29,"pricing":29,"motivation":29,"time_to_build":29,"stack":37,"openness":44,"stage":44,"team_size":44},[38],{"value":39,"source":40,"evidence":41},"Python","maker",[42],{"id":43},"post","unknown",{"praise":46,"objections":55,"requests":56,"questions":61,"alternatives":66},[47],{"value":48,"source":49,"evidence":50},"Commenter welcomes shipping a pretrained model with the code, since trained char-RNNs are hard to find","commenter",[51],{"id":52,"by":53,"text":54},"14966176","neoncontrails","Nice to see a preexisting model ship with the code. Always surprised me how difficult it is to find and compare trained RNNs, given how tiny a typical char-RNN model is, and how…",[],[57],{"value":58,"source":49,"evidence":59},"Load different pretrained models on demand to compare parameters such as window size",[60],{"id":52,"by":53,"text":54},[62],{"value":63,"source":49,"evidence":64},"How do longer window sizes (n>1000) affect output quality and topic consistency",[65],{"id":52,"by":53,"text":54},[],{"comments":31,"answered":31,"key":68},[],{"commenters":15,"max_depth":15,"maker_replies":31,"first_comment_mins":70,"lifespan_hours":71},2436,41,[73],[52,31,35,74,53,31],"project\u002Fpositive\u002Fidea",[76,77,80,83],{"id":4,"t":12,"points":14,"comments":15},{"id":78,"t":79,"points":15,"comments":31},"14970991",1502296731,{"id":81,"t":82,"points":15,"comments":15},"16849298",1523889845,{"id":84,"t":85,"points":15,"comments":31},"16876519",1524150389,[87,92,97,102,106,111],{"id":88,"name":89,"points":90,"category":91},"14947361","Is the stock market going to crash?",813,"finance_crypto",{"id":93,"name":94,"points":95,"category":96},"14957201","unwttng.com",256,"learning_science",{"id":98,"name":99,"points":100,"category":101},"14950699","Eliot and Me Estimator",220,"web_saas",{"id":103,"name":104,"points":105,"category":20},"14973741","Transformation Invariant Reverse Image Search",166,{"id":107,"name":108,"points":109,"category":110},"14947449","AsciiDots",148,"languages",{"id":112,"name":113,"points":114,"category":115},"14919534","RESTful Doom",136,"games",13,false,"v2",{"id":120,"name":5,"line":6,"ai":117,"cat":20,"year":13,"points":14,"comments":15,"rec":28,"s":29,"top":18,"kw":121,"maker":11,"domain":10,"tier":19,"attempts":122,"strip":123,"url":9,"mix":124,"praise":48,"objection":-1,"sub":21},14947671,[24,21,25,26,27],4,"5",[15,31,31,31],1794036557965]