[{"data":1,"prerenderedAt":431},["ShallowReactive",2],{"project-26048800":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":27,"sentiment":28,"stance_mix":29,"scored_comments":34,"aspects":35,"strip":45,"post_text":46,"what":47,"said":90,"maker_in_thread":223,"shape":230,"grid":234,"attempts":394,"same_week":395,"maker_shows":37,"maker_hidden":423,"reader_version":424,"card":425},"26048800","LibreTranslate","Open-source neural machine translation API","title","Show HN: LibreTranslate – Open-source neural machine translation API","https:\u002F\u002Flibretranslate.com\u002F","libretranslate.com","pjfin123",1612637322,2021,216,65,205,2533,9,"notable","languages","machine translation","api",[21,24,25,26],"open source translation","translation api","neural translation","mixed",-0.183,{"positive":30,"neutral":31,"mixed":32,"critical":33},4,12,5,10,31,{"idea":36,"performance":39,"design":42,"execution":44},{"positive":37,"critical":38},2,3,{"positive":40,"critical":41},1,7,{"positive":40,"critical":43},0,{"positive":43,"critical":43},[32,40,43,43,40,43,38,38,43,38,38,38,38,38,43,38,43,32,38,43,40,32,38,40,43,43,43,43,32,40,38,38,43,43,43,38,40,38,40,40],null,{"problem":48,"audience":60,"pricing":65,"motivation":46,"time_to_build":46,"stack":70,"openness":87,"stage":88,"team_size":89},{"value":49,"source":50,"evidence":51},"Free, open-source machine translation between languages","inferred",[52,56],{"id":53,"by":54,"text":55},"26051849","franciscop","This is pretty sweet! Specially it being free. I just added support for Libre Translate to my Javascript translation library, `translate` [1]: ``` import translate from…",{"id":57,"by":58,"text":59},"26049993","Mizza","Very glad to see something like this, this was on my high-priority Free software needs list. I'd very much like if it could be used programmatically, and not just via API -…",{"value":61,"source":50,"evidence":62},"Developers wanting free translation in their software",[63,64],{"id":53,"by":54,"text":55},{"id":57,"by":58,"text":59},{"value":66,"source":67,"evidence":68},"free","commenter",[69],{"id":53,"by":54,"text":55},[71,81],{"value":72,"source":73,"evidence":74},"Argos Translate (OpenNMT-based Transformer models)","maker",[75,78],{"id":76,"by":11,"text":77},"26051457","Interesting, thanks for the detailed write up. > The translation of the first sentence conveys the meaning but changes the date from February 7 to October 7 and drops the year.…",{"id":79,"by":11,"text":80},"26049475","The process for translation is to \"tokenize\" a stream of characters into \"tokens\" like this: \"Hello Sarah, what time is it?\" \u003CHello>\u003C_Sarah>\u003C,>\u003C_what>\u003C_time>\u003C_is>\u003C_it>\u003C?> Then…",{"value":82,"source":73,"evidence":83},"OPUS parallel corpus training data",[84],{"id":85,"by":11,"text":86},"26049521","OPUS parallel corpus: http:\u002F\u002Fopus.nlpl.eu\u002F It's really great, they have a large amount of data and organize it to make it easy to access.","open_source","launched","unknown",{"praise":91,"objections":111,"requests":148,"questions":168,"alternatives":187},[92,101],{"value":93,"source":67,"evidence":94},"Free and open source, a welcome alternative to paid translation APIs",[95,96,97],{"id":53,"by":54,"text":55},{"id":57,"by":58,"text":59},{"id":98,"by":99,"text":100},"26096911","dwheeler","An interesting alternative OSS natural language translator is Apertium: https:\u002F\u002Fwww.apertium.org\u002F Their public site doesn't support English\u002FFrench, though, so it's hard for me to…",{"value":102,"source":67,"evidence":103},"Decent quality for many languages and a clean UI",[104,108],{"id":105,"by":106,"text":107},"26049460","yamrzou","Well done. The UI is nice and easy to use. The results looked good for the few sentences I tried (Arabic \u003C--> English). May I know which datasets were used to train the models?",{"id":109,"by":99,"text":110},"26092524","This is really cool. I sent it several paragraphs to translate and the results were quite understandable. On a quick check I don't think the quality is as good as Google…",[112,127,141],{"value":113,"source":67,"evidence":114},"Translation quality is weak versus DeepL and Google, especially for Chinese",[115,119,123],{"id":116,"by":117,"text":118},"26051687","eruci","Actually DeepL seems to be the best of the bunch. Here is another test: Input: 北京市西城区地安门西大街49号 DeepL: No.49, Di'anmen West Street, Xicheng District, Beijing Google: 49 Di'anmen…",{"id":120,"by":121,"text":122},"26054459","guilamu","Deepl is the best in 99% of the case. I have no idea how they managed to get so much better than other players in the translating game.",{"id":124,"by":125,"text":126},"26062744","Ambix","Pretty interesting but the quality is far from top notch. Most reasonable and 20x cheaper alternative to Google Translate as I know is [1] [1] https:\u002F\u002Fwww.deeptranslate.net\u002F",{"value":128,"source":67,"evidence":129},"Frequent errors on addresses, dates and grammar",[130,134,137],{"id":131,"by":132,"text":133},"26051318","tom_mellior","I tried this on a fragment of https:\u002F\u002Fde.wikipedia.org\u002Fwiki\u002F 's Artikel des Tages (article of the day) box for today (February 7, 2021). Input: \"Das Frauen­stimm­recht in der…",{"id":135,"by":117,"text":136},"26051096","Comparing this to IBM Watson and Google Translate: Chinese Input: 古荡塘苗路华星现代产业园E座正门 LibreTranslate: Ordinary gate of the modern industrial plant of the Hyong Chung Chung IBM…",{"id":138,"by":139,"text":140},"26054180","mhkool","“He didn’t know what to think” is translated to Portuguese “Ele não sabia o que fazer” which is incorrect. Fazer is “to do” not “to think”",{"value":142,"source":67,"evidence":143},"Rate limit errors",[144],{"id":145,"by":146,"text":147},"26049706","otagekki","I had tried translating a test sentence and I got a rate limit related error... Would be glad to see this for Malagasy though",[149,157,161],{"value":150,"source":67,"evidence":151},"Support more languages such as Cantonese and Malagasy",[152,156],{"id":153,"by":154,"text":155},"26050589","fartcannon","Good. Now can we finally stop pretending Cantonese doesn't exist?",{"id":145,"by":146,"text":147},{"value":158,"source":67,"evidence":159},"Programmatic bindings for C, Python, Rust",[160],{"id":57,"by":58,"text":59},{"value":162,"source":67,"evidence":163},"Better handling of capitalization and short phrases",[164],{"id":165,"by":166,"text":167},"26049635","bmicraft","Well it would be nice if it worked, but it could not even translate \"merry christmas\" to German, just left it as is. Apparently it needs the C to be capitalized ...",[169,173,180],{"value":170,"source":67,"evidence":171},"Which datasets were used to train the models",[172],{"id":105,"by":106,"text":107},{"value":174,"source":67,"evidence":175},"Why small input changes alter output",[176],{"id":177,"by":178,"text":179},"26049387","jarym","Interesting if I do English to French on the following: Hello Sarah, what time is it? it translates to Bonjour Sarah, quelle heure est-il ? Now if I change the input to Hello…",{"value":181,"source":67,"evidence":182},"Whether the model keeps learning",[183],{"id":184,"by":185,"text":186},"26053520","LockAndLol","Does it continue learning? It might be good to somehow federalize it in order for instances to share knowledge or results. Were its current knowledge to stay static, I imagine it…",[188,193,201,205,212,219],{"value":189,"source":67,"evidence":190},"Google Translate",[191,192],{"id":135,"by":117,"text":136},{"id":109,"by":99,"text":110},{"value":194,"source":67,"evidence":195},"DeepL",[196,197],{"id":116,"by":117,"text":118},{"id":198,"by":199,"text":200},"26052335","zmix","This could become big, however, there is still work to do, in order to match something like DeepL[1] I took [2] and had it translated on both sites, and while the translation was…",{"value":202,"source":67,"evidence":203},"IBM Watson Language Translator",[204],{"id":135,"by":117,"text":136},{"value":206,"source":67,"evidence":207},"Apertium",[208,209],{"id":98,"by":99,"text":100},{"id":210,"by":11,"text":211},"26051424","It was partially trained on EU and UN documents. Interesting observation, less emphasis on parliamentary\u002Flegal data vs a data set like OpenSubtitles, which has more casual…",{"value":213,"source":67,"evidence":214},"ModernMT",[215],{"id":216,"by":217,"text":218},"26053611","ai_ja_nai","There is plenty of FOSS powered services like this out there, for example https:\u002F\u002Fwww.modernmt.com\u002Ftranslate\u002F: it is a EU funded project. What I dislike here is that it looks…",{"value":220,"source":67,"evidence":221},"DeepTranslate",[222],{"id":124,"by":125,"text":126},{"comments":33,"answered":32,"key":224},[225,228,229],{"id":226,"text":227},"26051911","Very cool, thanks or making this!",{"id":210,"text":211},{"id":76,"text":77},{"commenters":231,"max_depth":41,"maker_replies":33,"first_comment_mins":232,"lifespan_hours":233},41,71,102,[235,237,239,241,245,248,251,254,255,256,260,263,266,270,274,276,279,280,283,285,287,288,291,293,295,298,299,301,304,306,310,313,316,318,319,323,325,327,329,331,334,337,339,341,344,346,348,350,352,353,356,358,360,362,364,366,368,372,374,376,378,381,383,387,390],[53,43,32,236,54,43],"project\u002Fpositive\u002Fidea",[226,43,43,238,11,40],"maker\u002Fpositive\u002Fnone",[135,40,40,240,117,43],"project\u002Fcritical\u002Fperformance",[242,40,43,243,244,43],"26051167","commenter\u002Fneutral\u002Fnone","cdrini",[246,40,43,243,247,43],"26051203","parsecs",[249,40,43,243,250,43],"26051371","rococode",[252,40,43,243,253,43],"26051458","yorwba",[116,40,40,240,117,43],[120,40,43,243,121,43],[257,40,38,258,259,43],"26065162","project\u002Fneutral\u002Fperformance","jimz",[261,40,38,258,262,43],"26051196","elisaado",[264,40,43,243,265,43],"26052648","nerdponx",[267,40,38,268,269,43],"26051220","project\u002Fneutral\u002Fidea","iujjkfjdkkdkf",[271,40,43,272,273,43],"26069718","other\u002Fneutral\u002Fnone","michaelbrave",[198,37,38,275,199,43],"project\u002Fmixed\u002Fperformance",[277,37,43,243,278,43],"26053669","An0mammall",[131,38,38,275,132,43],[281,38,38,268,282,43],"26051389","ascorbic",[210,38,43,284,11,40],"maker\u002Fneutral\u002Fidea",[286,38,38,258,132,43],"26051493",[76,38,43,238,11,40],[289,30,38,275,290,43],"26051973","grishka",[292,30,43,243,132,43],"26053907",[294,30,43,243,290,43],"26056812",[296,30,43,243,297,43],"26055664","zoomablemind",[177,32,38,258,178,43],[79,32,43,300,11,40],"maker\u002Fneutral\u002Fperformance",[302,32,43,243,303,43],"26049481","tejtm",[57,305,32,236,58,43],6,[307,305,38,308,309,43],"26050078","project\u002Fneutral\u002Fnone","rahimnathwani",[311,305,43,312,11,40],"26050115","maker\u002Fneutral\u002Fnone",[314,305,43,243,315,43],"26050462","drusepth",[317,305,43,272,58,43],"26051133",[165,41,40,240,166,43],[320,41,32,321,322,43],"26049765","project\u002Fpositive\u002Fperformance","dheera",[324,41,43,300,11,40],"26049878",[109,326,38,275,99,43],8,[153,18,40,328,154,43],"project\u002Fcritical\u002Fidea",[330,18,43,243,253,43],"26051151",[332,18,43,243,333,43],"26051231","whimsicalism",[335,18,43,243,336,43],"26051594","siver_john",[338,18,43,243,253,43],"26054005",[340,18,43,243,259,43],"26065779",[342,18,43,243,343,43],"26055466","ficklepickle",[345,18,43,243,253,43],"26057656",[105,33,32,347,106,43],"project\u002Fpositive\u002Fdesign",[85,33,43,349,11,40],"maker\u002Fneutral\u002Fdocs",[216,351,40,328,217,43],11,[98,31,38,308,99,43],[354,355,38,258,253,43],"26049393",13,[357,355,43,300,11,40],"26049558",[359,355,38,258,253,43],"26050149",[361,355,43,300,11,40],"26050609",[363,355,43,243,253,43],"26051087",[365,355,43,312,11,40],"26051152",[367,355,43,243,253,43],"26051351",[369,370,38,268,371,43],"26059646",14,"rini17",[184,373,38,268,185,43],15,[124,375,40,240,125,43],16,[138,377,40,240,139,43],17,[145,379,38,380,146,43],18,"project\u002Fmixed\u002Fexecution",[382,379,40,328,253,43],"26050258",[384,385,40,240,386,43],"26054579",19,"Aeolun",[388,385,43,243,389,43],"26054678","kome",[391,392,40,240,393,43],"26053664",20,"konart",[],[396,401,406,411,415,419],{"id":397,"name":398,"points":399,"category":400},"26069621","Clerk",582,"web_saas",{"id":402,"name":403,"points":404,"category":405},"26087892","Building Your Mouseless Development Environment",353,"learning_science",{"id":407,"name":408,"points":409,"category":410},"26042463","Ht",251,"dev_tools",{"id":412,"name":413,"points":414,"category":410},"26048993","Epub.to",246,{"id":416,"name":417,"points":418,"category":400},"26038501","Remarkbox",233,{"id":420,"name":421,"points":422,"category":410},"26056052","Stamp",195,false,"v2",{"id":426,"name":5,"line":6,"ai":423,"cat":20,"year":13,"points":14,"comments":15,"rec":27,"s":427,"top":18,"kw":428,"maker":11,"domain":10,"tier":19,"attempts":43,"strip":429,"url":9,"mix":430,"praise":93,"objection":113,"sub":21},26048800,-0.18,[21,24,25,26],"5100103303333303053015310000513300031311",[30,31,32,33],1794036602708]