[{"data":1,"prerenderedAt":196},["ShallowReactive",2],{"project-12239096":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":29,"sentiment":30,"stance_mix":31,"scored_comments":33,"aspects":34,"strip":37,"post_text":30,"what":38,"said":54,"maker_in_thread":121,"shape":127,"grid":132,"attempts":159,"same_week":160,"maker_shows":122,"maker_hidden":189,"reader_version":190,"card":191},"12239096","deep learning models","Set of trained deep learning models for computer vision","title","Show HN: Set of trained deep learning models for computer vision","https:\u002F\u002Fgithub.com\u002Ffchollet\u002Fdeep-learning-models","github.com","fchollet",1470506617,2016,127,15,281,2264,13,"notable","ai","pretrained computer vision models","library",[24,25,26,27,28],"pretrained models","keras","image classification","computer vision","model weights","quiet",null,{"positive":32,"neutral":33,"mixed":32,"critical":32},0,3,{"idea":35,"docs":36},{"positive":32,"critical":32},{"positive":32,"critical":32},[32,33,32,33,32,32,32,32,33,32,32,32,32,32,32],{"problem":30,"audience":30,"pricing":30,"motivation":30,"time_to_build":30,"stack":39,"openness":52,"stage":53,"team_size":53},[40,48],{"value":41,"source":42,"evidence":43},"Keras","inferred",[44],{"id":45,"by":46,"text":47},"12241595","nl","Yes - pretty much every Deep Learning library. Caffe, Torch, Theano, TensorFlow etc (that's kinda what this link is about?) Just use Keras on top of TensorFlow as shown at this…",{"value":49,"source":42,"evidence":50},"TensorFlow",[51],{"id":45,"by":46,"text":47},"open_source","unknown",{"praise":55,"objections":56,"requests":73,"questions":101,"alternatives":106},[],[57],{"value":58,"source":59,"evidence":60},"Licensing of the model weights is unclear","commenter",[61,65,69],{"id":62,"by":63,"text":64},"12239305","minimaxir","Are you allowed to redistribute the models under the MIT License?",{"id":66,"by":67,"text":68},"12240714","ma2rten","Don't take my word for it, but actually as far as I know both in the US and EU data (including model weights) can't be copyrighted.",{"id":70,"by":71,"text":72},"12241270","cmrx64","Even if this might be true, including a license of some kind is really a best practice at this point. Something like a Creative Commons license is probably more appropriate than…",[74,81,91],{"value":75,"source":59,"evidence":76},"Clearer licensing information for the weights",[77,78],{"id":70,"by":71,"text":72},{"id":79,"by":11,"text":80},"12239338","The code is under the MIT license, not the weights. The weights are under their respective licenses. The weights are not included in the git tree and are thus not covered by the…",{"value":82,"source":59,"evidence":83},"Support for fine-tuning on new image datasets",[84,88],{"id":85,"by":86,"text":87},"12239568","Omnipresent","Can the classify images example be modified to train on other (new) images. For example, images of screenshots to identify elements that are in the images (such as: word…",{"id":89,"by":11,"text":90},"12239591","Yes, you can use these models for fine-tuning (or feature extraction) on a new dataset. This tutorial would be a good place to start (esp. sections 2 and 3):…",{"value":92,"source":59,"evidence":93},"A registry for loading and sharing models",[94,98],{"id":95,"by":96,"text":97},"12240390","chris_va","Question: Does anyone know of a library for loading models\u002Fweights from a registry of some sort?",{"id":99,"by":96,"text":100},"12242345","Well, this is fairly manual. More like: my_model = registry.get(\"tensorflow:\u002F\u002Fgithub.com\u002Fasdf\u002Fmodels\u002Fimagine\u002Flatest\") ... my_model.push(\"...\")",[102],{"value":103,"source":59,"evidence":104},"Can the models be redistributed under MIT?",[105],{"id":62,"by":63,"text":64},[107,114],{"value":108,"source":59,"evidence":109},"Caffe Model Zoo",[110],{"id":111,"by":112,"text":113},"12239776","terhechte","Caffenet also offers a set of pre-trained models in their \"model zoo\": https:\u002F\u002Fgithub.com\u002FBVLC\u002Fcaffe\u002Fwiki\u002FModel-Zoo",{"value":115,"source":59,"evidence":116},"spaCy Sputnik",[117],{"id":118,"by":119,"text":120},"12242159","elyase","You mean something like this: https:\u002F\u002Fgithub.com\u002Fspacy-io\u002Fsputnik ?",{"comments":122,"answered":123,"key":124},2,1,[125,126],{"id":89,"text":90},{"id":79,"text":80},{"commenters":128,"max_depth":129,"maker_replies":122,"first_comment_mins":130,"lifespan_hours":131},11,5,35,25,[133,135,137,139,140,141,143,146,149,150,152,153,154,156,158],[111,32,32,134,112,32],"other\u002Fneutral\u002Fnone",[85,123,33,136,86,32],"project\u002Fneutral\u002Fidea",[89,123,32,138,11,123],"project\u002Fneutral\u002Fdocs",[62,122,33,136,63,32],[79,122,32,138,11,123],[66,122,32,142,67,32],"commenter\u002Fneutral\u002Fidea",[144,122,32,142,145,32],"12240730","akhilcacharya",[147,122,32,134,148,32],"12241204","praccu",[70,122,33,138,71,32],[95,33,32,151,96,32],"other\u002Fneutral\u002Fidea",[45,33,32,134,46,32],[99,33,32,151,96,32],[155,33,32,142,46,32],"12242705",[157,33,32,142,96,32],"12243176",[118,33,32,134,119,32],[],[161,166,171,175,180,184],{"id":162,"name":163,"points":164,"category":165},"12260794","Generating fantasy maps",1004,"creative",{"id":167,"name":168,"points":169,"category":170},"12260958","Xaddress",415,"languages",{"id":172,"name":173,"points":174,"category":170},"12227507","asyncpg",296,{"id":176,"name":177,"points":178,"category":179},"12240542","Gping.io",184,"hardware",{"id":181,"name":182,"points":183,"category":170},"12241332","cfilter",181,{"id":185,"name":186,"points":187,"category":188},"12232293","Web Dev with Golang Anti-Textbook",158,"learning_science",false,"v2",{"id":192,"name":5,"line":6,"ai":189,"cat":20,"year":13,"points":14,"comments":15,"rec":29,"s":30,"top":18,"kw":193,"maker":11,"domain":10,"tier":19,"attempts":32,"strip":194,"url":9,"mix":195,"praise":-1,"objection":58,"sub":21},12239096,[24,25,26,27,28],"030300003000000",[32,33,32,32],1794036546525]