[{"data":1,"prerenderedAt":187},["ShallowReactive",2],{"project-17294875":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":15,"aspects":35,"strip":39,"post_text":41,"what":42,"said":75,"maker_in_thread":130,"shape":132,"grid":135,"attempts":148,"same_week":149,"maker_shows":178,"maker_hidden":179,"reader_version":180,"card":181},"17294875","ModelDepot Percept","Deploy and Retrain Cutting Edge ML Image Recognition via REST API","title","Show HN: Deploy and Retrain Cutting Edge ML Image Recognition via REST API","https:\u002F\u002Fmodeldepot.io\u002Fpercept","modeldepot.io","hsikka",1528820952,2018,13,7,1485,2155,69,"shown","ai","ml model deployment","api",[24,25,26,27,28],"image classification","ml deployment","rest api","model retraining","transfer learning","mixed",-0.267,{"positive":32,"neutral":33,"mixed":34,"critical":33},1,3,0,{"idea":36,"pricing":37},{"positive":32,"critical":32},{"positive":34,"critical":38},2,[40,32,33,32,33,32,33],5,null,{"problem":43,"audience":51,"pricing":58,"motivation":68,"time_to_build":41,"stack":72,"openness":73,"stage":74,"team_size":73},{"value":44,"source":45,"evidence":46},"Deploying and retraining ML image classification models is hard for everyday developers","commenter",[47],{"id":48,"by":49,"text":50},"17295876","mikeshi42","Hey Everyone! Harsh and I built ModelDepot with the goal of empowering everyday developers to use Machine Learning simply and quickly. After having thousands of engineers use…",{"value":52,"source":45,"evidence":53},"Everyday developers and engineering teams without deep ML expertise",[54,55],{"id":48,"by":49,"text":50},{"id":56,"by":49,"text":57},"17295709","Hey! I think all of those points are 100% valid and I understand where you're coming from. I'd like to preface that the product you see today isn't the end goal we're trying to…",{"value":59,"source":45,"evidence":60},"Pay-as-you-go usage pricing per API request, with free credits offered",[61,64,67],{"id":62,"by":49,"text":63},"17296094","The pricing model is to help support our continual development costs of the platform, it definitely takes a bit of time to develop the tech, and continually test it with our…",{"id":65,"by":49,"text":66},"17296498","I don't think usage based pricing is esoteric, if you look towards products in the enterprise space (Gitlab, Splunk, Mongo, etc.) They're all based off of a usage metric of some…",{"id":48,"by":49,"text":50},{"value":69,"source":45,"evidence":70},"Empower everyday developers to use machine learning simply and quickly",[71],{"id":48,"by":49,"text":50},[],"unknown","launched",{"praise":76,"objections":77,"requests":103,"questions":112,"alternatives":121},[],[78,88,95,99],{"value":79,"source":45,"evidence":80},"Usage-based pricing is misplaced for a self-hosted product",[81,85],{"id":82,"by":83,"text":84},"17296371","chatmasta","Usage based pricing seems very misplaced for a self hosted solution. Questions of enforceability aside (all I need to do is remove the telemetry code to get it free), the issue…",{"id":86,"by":83,"text":87},"17295893","How do you enforce request limits when the code is not hosted on your server? And why should I pay tiered based on number of requests if I’m the one paying the infrastructure…",{"value":89,"source":45,"evidence":90},"Real ML work is data cleaning and model evaluation, which the product does not cover",[91],{"id":92,"by":93,"text":94},"17295057","mlthoughts2018","I really think these types of services miss the big picture. The hard part of creating these services in-house is not wrapping things up in a single discrete unit of deployment…",{"value":96,"source":45,"evidence":97},"Usage metering can be bypassed by removing telemetry code",[98],{"id":82,"by":83,"text":84},{"value":100,"source":45,"evidence":101},"Vendor lock-in and dependence on in-house ML expertise",[102],{"id":92,"by":93,"text":94},[104,108],{"value":105,"source":45,"evidence":106},"Offer a licensing fee option for self-hosted deployment",[107],{"id":82,"by":83,"text":84},{"value":109,"source":45,"evidence":110},"Explain how request limits are enforced when code runs off-server",[111],{"id":86,"by":83,"text":87},[113,117],{"value":114,"source":45,"evidence":115},"How are requests enforced when hosting is not on the maker's server",[116],{"id":86,"by":83,"text":87},{"value":118,"source":45,"evidence":119},"What latency and throughput numbers apply at scale",[120],{"id":92,"by":93,"text":94},[122,126],{"value":123,"source":45,"evidence":124},"Google AutoML",[125],{"id":92,"by":93,"text":94},{"value":127,"source":45,"evidence":128},"AWS Rekognition",[129],{"id":92,"by":93,"text":94},{"comments":34,"answered":34,"key":131},[],{"commenters":33,"max_depth":133,"maker_replies":34,"first_comment_mins":134,"lifespan_hours":38},6,15,[136,138,140,142,144,146,147],[48,34,40,137,49,34],"project\u002Fpositive\u002Fidea",[92,32,32,139,93,34],"project\u002Fcritical\u002Fidea",[56,32,33,141,49,34],"maker\u002Fneutral\u002Fidea",[86,32,32,143,83,34],"project\u002Fcritical\u002Fpricing",[62,32,33,145,49,34],"project\u002Fneutral\u002Fpricing",[82,32,32,143,83,34],[65,32,33,145,49,34],[],[150,155,159,164,169,173],{"id":151,"name":152,"points":153,"category":154},"17286770","Ancient Earth",384,"learning_science",{"id":156,"name":157,"points":158,"category":154},"17272516","Scaler (grunfy)",277,{"id":160,"name":161,"points":162,"category":163},"17291269","Dilay",227,"creative",{"id":165,"name":166,"points":167,"category":168},"17294548","Fo",165,"languages",{"id":170,"name":171,"points":172,"category":168},"17255418","µFSM",154,{"id":174,"name":175,"points":176,"category":177},"17289801","T2b",117,"dev_tools",4,false,"v2",{"id":182,"name":5,"line":6,"ai":179,"cat":20,"year":13,"points":14,"comments":15,"rec":29,"s":183,"top":18,"kw":184,"maker":11,"domain":10,"tier":19,"attempts":34,"strip":185,"url":9,"mix":186,"praise":-1,"objection":79,"sub":21},17294875,-0.27,[24,25,26,27,28],"5131313",[32,33,34,33],1794036566908]