[{"data":1,"prerenderedAt":332},["ShallowReactive",2],{"project-35649496":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":36,"aspects":37,"strip":44,"post_text":46,"what":47,"said":72,"maker_in_thread":143,"shape":155,"grid":159,"attempts":292,"same_week":293,"maker_shows":157,"maker_hidden":324,"reader_version":325,"card":326},"35649496","WasmGPT","“ChatGPT” in the browser, no WebGPU and no server needed","title","Show HN: WasmGPT – “ChatGPT” in the browser, no WebGPU and no server needed","https:\u002F\u002Flxe.co\u002Fwasmgpt\u002F","lxe.co","lxe",1682047483,2023,185,50,310,3109,10,"notable","ai","local llm chat","web_app",[24,25,26,27,28],"wasm","browser llm","webassembly","offline chatbot","ggml","roasted",-0.422,{"positive":32,"neutral":33,"mixed":34,"critical":35},3,8,2,17,30,{"performance":38,"execution":40,"idea":42,"design":43},{"positive":39,"critical":35},1,{"positive":39,"critical":41},0,{"positive":39,"critical":41},{"positive":41,"critical":41},[39,41,39,41,39,41,41,39,45,41,39,39,41,41,32,39,41,39,39,32,32,39,39,32,41,39,45,39,39,41,39,39,32,45,41,39,41,32,32,32],5,"Using threaded emscripten to speed up the generation and offload the main loop. No SIMD or other optimizations. Might work faster with #enable-experimental-webassembly-features enabled.\n\nTested in x86 Chrome and Firefox, Apple Silicon Safari\n\nRun it yourself: https:\u002F\u002Fgithub.com\u002Flxe\u002Fggml\u002Ftree\u002Fwasm-demo\n\nThanks, https:\u002F\u002Fgithub.com\u002Fggerganov\u002Fggml,",{"problem":48,"audience":54,"pricing":54,"motivation":54,"time_to_build":54,"stack":55,"openness":69,"stage":70,"team_size":71},{"value":49,"source":50,"evidence":51},"Runs a chat language model in the browser without a server or WebGPU","maker",[52],{"id":53},"post",null,[56,60,63],{"value":57,"source":50,"evidence":58},"WebAssembly via threaded emscripten",[59],{"id":53},{"value":28,"source":50,"evidence":61},[62],{"id":53},{"value":64,"source":50,"evidence":65},"Cerebras-GPT 1.3B finetuned with LoRA on Alpaca",[66],{"id":67,"by":11,"text":68},"35650124","I'm using https:\u002F\u002Fhuggingface.co\u002Flxe\u002FCerebras-GPT-1.3B-Alpaca-SP which is https:\u002F\u002Fhuggingface.co\u002Fcerebras\u002FCerebras-GPT-1.3B finetuned using LoRA on Alpaca dataset. The cerebras…","open_source","prototype","unknown",{"praise":73,"objections":86,"requests":107,"questions":123,"alternatives":131},[74],{"value":75,"source":76,"evidence":77},"Fun to try and impressive that it runs locally in the browser","commenter",[78,82],{"id":79,"by":80,"text":81},"35649553","throwaway888abc","Impressive! Just tried and woks fine (linux,i5,chrome)",{"id":83,"by":84,"text":85},"35652768","jijji","regardless of gibberish or not it's pretty amazing that you can put it in the browser as Wasm, locally inferred",[87,101],{"value":88,"source":76,"evidence":89},"Output quality is gibberish or nonsensical for most questions",[90,93,97],{"id":91,"by":84,"text":92},"35650072","I tested this using about 10 different questions that were answered correctly on chatGPT, all of the answers produced gibberish... is it because you're using the GPT-2 model?",{"id":94,"by":95,"text":96},"35650996","hathym","It failed to accomplish all the trivial tasks I've requested, rendering it unusable for the time being. ``` User: implement a python function that takes two numbers and return…",{"id":98,"by":99,"text":100},"35650402","nwbort","Probably having too much fun with this, thanks ``` Human: What is 3+3 Assistant: 3+3 Human: What does it equal Assistant: It is equal to the following: Assistant: It is equal to…",{"value":102,"source":50,"evidence":103},"Much slower than WebGPU alternatives",[104],{"id":105,"by":11,"text":106},"35649981","It is not. It's much slower. But it doesn't require you to use a canary browser with a command line flag, so that's that. EDIT: I guess I should give credit here...…",[108,119],{"value":109,"source":76,"evidence":110},"Progress indicator while the answer generates",[111,115],{"id":112,"by":113,"text":114},"35651946","Lex-2008","Nice stuff! A bit of feedback, if I may: * maybe it worth showing some feedback when AI is thinking\u002Fcomposing the answer vs. when it's already finished. Currently, on my machine…",{"id":116,"by":117,"text":118},"35652389","pmarreck","A progress indicator on the download might’ve been nice",{"value":120,"source":76,"evidence":121},"Fix non-Latin and UTF-8 input handling",[122],{"id":112,"by":113,"text":114},[124],{"value":125,"source":76,"evidence":126},"Is it faster than WebGPU and why",[127],{"id":128,"by":129,"text":130},"35649969","eachro","Is it faster than webgpu? If so, what makes it so?",[132],{"value":133,"source":76,"evidence":134},"ChatGPT",[135,139],{"id":136,"by":137,"text":138},"35651070","teekert",":D, good and funny to read indicator of the models \"intelligence\". It's at least more fun than ChatGPT (I am feeling good and so have I!): User: Hi wasmGPT, how are you?…",{"id":140,"by":141,"text":142},"35663255","terrycody","Sorry for a layman like me, what does this differ with ChatGPT web interface on openAI website? I can't get it what it does.",{"comments":144,"answered":32,"key":145},12,[146,149,152],{"id":147,"text":148},"35650840","Keep in mind that this demo doesn't have context memory to keep it performant. Every new question starts without the knowledge of previous answers.",{"id":150,"text":151},"35651162","Thank you for your amazing work! This is inspired by your whisper wasm demo! I can barely keep up with the changes to your repos. You landed gpt-neox \u002F stablelm support right…",{"id":153,"text":154},"35657124","This one is my favorite!",{"commenters":156,"max_depth":157,"maker_replies":144,"first_comment_mins":33,"lifespan_hours":158},34,4,108,[160,164,166,167,171,174,177,180,184,187,190,194,196,198,199,200,202,205,207,209,213,216,219,223,227,229,232,234,236,238,240,243,246,249,252,256,258,259,262,265,267,271,273,276,279,281,282,284,287,289],[161,41,39,162,163,41],"35650823","project\u002Fcritical\u002Fperformance","BHSPitMonkey",[147,41,41,165,11,39],"project\u002Fneutral\u002Fperformance",[136,41,39,162,137,41],[168,41,41,169,170,41],"35651607","other\u002Fneutral\u002Fnone","moffkalast",[172,41,41,169,173,41],"35669453","thrill",[175,41,39,162,176,41],"35653389","sheepscreek",[178,41,41,169,179,41],"35653577","RecycledEle",[181,41,41,182,183,41],"35653751","commenter\u002Fneutral\u002Fnone","tempaccount420",[185,41,39,162,186,41],"35689452","taneq",[188,39,32,165,189,41],"35651488","Gnarl",[191,34,45,192,193,41],"35651008","project\u002Fpositive\u002Fperformance","ggerganov",[150,34,41,195,11,39],"commenter\u002Fpositive\u002Fnone",[197,32,39,162,189,41],"35651459",[153,32,41,182,11,39],[98,157,39,162,99,41],[201,157,41,165,11,39],"35650509",[203,157,41,169,204,41],"35651425","mdp2021",[112,45,32,206,113,41],"project\u002Fmixed\u002Fexecution",[208,45,41,182,11,39],"35654880",[210,211,39,162,212,41],"35652477",6,"iam-TJ",[214,215,41,165,11,39],"35650143",7,[217,33,39,162,218,41],"35651709","fsniper",[220,221,39,162,222,41],"35702206",9,"temp2022account",[224,18,32,225,226,41],"35651431","project\u002Fmixed\u002Fperformance","NovaDudely",[228,18,41,182,11,39],"35651456",[230,18,32,165,231,41],"35652233","monkey_monkey",[94,233,39,162,95,41],11,[235,233,39,162,137,41],"35651053",[237,233,41,165,11,39],"35651197",[239,233,32,165,113,41],"35651793",[241,233,41,169,242,41],"35668557","musicale",[244,144,39,162,245,41],"35653465","axismundi",[79,247,45,248,80,41],13,"project\u002Fpositive\u002Fexecution",[250,247,41,251,11,39],"35649622","project\u002Fneutral\u002Fexecution",[253,254,39,162,255,41],"35652699",14,"c16",[91,257,39,162,84,41],15,[67,257,41,165,11,39],[260,257,39,162,261,41],"35650393","AyanRafique",[263,257,39,162,264,41],"35650456","isuckatcoding",[266,257,41,169,204,41],"35651458",[268,257,32,269,270,41],"35650508","project\u002Fneutral\u002Fnone","Vanit",[83,257,45,272,84,41],"project\u002Fpositive\u002Fidea",[274,257,41,182,275,41],"35657761","circuit10",[277,257,39,162,278,41],"35651297","kieckerjan",[128,280,32,165,129,41],16,[105,280,41,165,11,39],[140,280,32,283,141,41],"project\u002Fneutral\u002Fidea",[285,280,32,283,286,41],"35668401","3np",[288,35,41,165,11,39],"35650758",[116,290,32,291,117,41],18,"project\u002Fneutral\u002Fdesign",[],[294,299,304,309,314,319],{"id":295,"name":296,"points":297,"category":298},"35639810","Stretch15",550,"consumer_social",{"id":300,"name":301,"points":302,"category":303},"35665142","Lost at SQL",547,"learning_science",{"id":305,"name":306,"points":307,"category":308},"35712163","ScrapScript",503,"languages",{"id":310,"name":311,"points":312,"category":313},"35690753","Homemade rocketship treehouse",410,"hardware",{"id":315,"name":316,"points":317,"category":318},"35649245","newgrounds.com",267,"creative",{"id":320,"name":321,"points":322,"category":323},"35712238","GB Renewables Map",245,"data",false,"v2",{"id":327,"name":5,"line":6,"ai":324,"cat":20,"year":13,"points":14,"comments":15,"rec":29,"s":328,"top":18,"kw":329,"maker":11,"domain":10,"tier":19,"attempts":41,"strip":330,"url":9,"mix":331,"praise":75,"objection":88,"sub":21},35649496,-0.42,[24,25,26,27,28],"1010100150110031011331130151101135010333",[32,33,34,35],1794036634354]