[{"data":1,"prerenderedAt":437},["ShallowReactive",2],{"project-40857009":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":89,"maker_in_thread":212,"shape":223,"grid":228,"attempts":400,"same_week":401,"maker_shows":45,"maker_hidden":429,"reader_version":430,"card":431},"40857009","Pretzel","Adding Mistral Codestral and GPT-4o to Jupyter Notebooks","title","Show HN: Adding Mistral Codestral and GPT-4o to Jupyter Notebooks","https:\u002F\u002Fgithub.com\u002Fpretzelai\u002Fpretzelai\u002Fblob\u002Fmain\u002FREADME.md","github.com","prasoonds",1719930236,2024,269,74,262,3515,8,"legend","dev_tools","notebook ai assistant","desktop_app",[24,25,26,27,28],"jupyter","ai code completion","codestral","jupyter notebook","code generation","curious",0.009,{"positive":32,"neutral":33,"mixed":34,"critical":35},5,10,0,4,19,{"design":38,"execution":40,"idea":42,"performance":43},{"positive":39,"critical":34},2,{"positive":41,"critical":41},1,{"positive":41,"critical":39},{"positive":34,"critical":34},[34,34,34,34,34,32,34,34,45,34,34,45,34,45,45,34,32,34,34,45,45,34,34,41,45,34,34,34,32,34,34,34,41,45,34,41,45,45,34,32],3,"Hey HN! We’ve forked Jupyter Lab and added AI code generation features that feel native and have all the context about your notebook. You can see a demo video (2 min) here: https:\u002F\u002Fwww.tella.tv\u002Fvideo\u002Fclxt7ei4v00rr09i5gt1laop6\u002Fview\n\nTry a hosted version here: https:\u002F\u002Fpretzelai.app\n\nJupyter is by far the most used Data Science tool. Despite its popularity, it still lacks good code-generation extensions. The flagship AI extension jupyter-ai lags far behind in features and UX compared to modern AI code generation and understanding tools (like https:\u002F\u002Fwww.continue.dev and https:\u002F\u002Fwww.cursor.com). A",{"problem":48,"audience":54,"pricing":61,"motivation":65,"time_to_build":72,"stack":73,"openness":86,"stage":87,"team_size":88},{"value":49,"source":50,"evidence":51},"Jupyter lacks good AI code generation and autocomplete, and jupyter-ai lags in UX and features","maker",[52],{"id":53},"post",{"value":55,"source":50,"evidence":56},"Data scientists and analysts who use Jupyter notebooks",[57,58],{"id":53},{"id":59,"by":11,"text":60},"40858509","Agree with Daksh in the sibling comment. I think it's like you said - different people have different workflows and some might prefer using VSCode. IME though, most data…",{"value":62,"source":50,"evidence":63},"Free; hosted AI Server or own OpenAI API key",[64],{"id":53},{"value":66,"source":50,"evidence":67},"Jupyter lacks good AI code generation, and the team wants to build a modern literate programming tool",[68,69],{"id":53},{"id":70,"by":11,"text":71},"40857807","So we discuss this briefly on our FAQ but let me try to expand on it. Our goal is to make a modern literate programming tool. On a surface level, a tool like that would end up…",null,[74,78,82],{"value":75,"source":50,"evidence":76},"Jupyter Lab (fork)",[77],{"id":53},{"value":79,"source":50,"evidence":80},"Mistral Codestral",[81],{"id":53},{"value":83,"source":50,"evidence":84},"GPT-4o",[85],{"id":53},"open_source","launched","unknown",{"praise":90,"objections":106,"requests":151,"questions":166,"alternatives":182},[91,99],{"value":92,"source":93,"evidence":94},"Open-source approach praised over closed forks","commenter",[95],{"id":96,"by":97,"text":98},"40857666","TidbitsTornado","This is a great implementation by your team + contributors. Simple but effective. And nice to see you’ve kept it open source instead of some other Show HN submissions where they…",{"value":100,"source":93,"evidence":101},"Copilot-like features are useful in Jupyter",[102],{"id":103,"by":104,"text":105},"40857535","lschneider","Github Copilot is the most useful tool I've found in a long time and having that in Jupyter Notebooks is just awesome. I've been missing that for quite some time. Great work guys!",[107,122,133,144],{"value":108,"source":93,"evidence":109},"VSCode with Copilot or Cursor already covers this",[110,114,118],{"id":111,"by":112,"text":113},"40859157","chad1n","I don't really get the appeal of this, I'd just use vscode with Jupyter if I really wanted \"ai\" integration since I can then access the whole ecosystem of extensions. The idea…",{"id":115,"by":116,"text":117},"40859217","widepeepo8","Codeium(https:\u002F\u002Fcodeium.com\u002F) already supports this, along with VSCode jupyter notebook extensions. It has 400k downloads on the VSCode extension store. don't really see the…",{"id":119,"by":120,"text":121},"40857977","stared","https:\u002F\u002Fwww.cursor.com\u002F - an AI-first VS Code clone VS Code (and Cursor) has so nice Jupyter support that I find it much better to use it for my workflow, rather than using any…",{"value":123,"source":93,"evidence":124},"jupyter-ai and other existing extensions already cover this",[125,129],{"id":126,"by":127,"text":128},"40861416","f-lux","Have y'all seen Jupyter AI? Seems to do the same but better (more features, more mature codebase, better UI\u002FUX\u002FDX) while being a JupyterLab extension…",{"id":130,"by":131,"text":132},"40857940","williamstein","There are many other Jupyter notebooks with extensive AI integration. These are less (or not at all) open source, but more mature in some ways, having been iterated on for over a…",{"value":134,"source":93,"evidence":135},"Fork instead of extension is questioned",[136,140],{"id":137,"by":138,"text":139},"40857654","carreau","Curious about the limitations that made you fork it instead of making an extension.",{"id":141,"by":142,"text":143},"40863664","trungld","your work is basically a Jupyter extension (https:\u002F\u002Fgithub.com\u002Fpretzelai\u002Fpretzelai\u002Ftree\u002Fmain\u002Fpackages\u002Fpretzelai-extension). Why don't just create an extension like others in here…",{"value":145,"source":93,"evidence":146},"Installation hangs on default port 8888",[147],{"id":148,"by":149,"text":150},"40861716","downrightmike","I installed it into my jupyter env and it runs fine on port 8889, but the default port 8888, it just sits waiting on the AI to generate a response.",[152,159],{"value":153,"source":93,"evidence":154},"Support for local LLMs and Claude Sonnet 3.5",[155],{"id":156,"by":157,"text":158},"40857256","ramonverse","Ramon here, the other cofounder of Pretzel! Quick update: Based on some early feedback, we're already working on adding support for local LLMs and Claude Sonnet 3.5. Happy to…",{"value":160,"source":93,"evidence":161},"Note-taking tool with AI for general notes",[162],{"id":163,"by":164,"text":165},"40859134","morsch","These editors all focus on programming, does anybody have a recommendation for more general note-taking? I'd like to do things like organizing very rough notes, having them…",[167,175],{"value":168,"source":93,"evidence":169},"Why fork Jupyter instead of building an extension",[170,171],{"id":137,"by":138,"text":139},{"id":172,"by":173,"text":174},"40863467","mitjam","Are there technical reasons for the fork or could Pretzel have been implemented as a set of extensions?",{"value":176,"source":93,"evidence":177},"Why Codestral for autocomplete",[178],{"id":179,"by":180,"text":181},"40858213","dakshgupta","Curious on why you went with Codestral for autocomplete, does it outperform other local models? How is the performance compared to GPT or Claude for autocomplete? Any plans to…",[183,188,193,197,201],{"value":184,"source":93,"evidence":185},"VSCode with Copilot or Cursor",[186,187],{"id":119,"by":120,"text":121},{"id":115,"by":116,"text":117},{"value":189,"source":93,"evidence":190},"jupyter-ai",[191,192],{"id":130,"by":131,"text":132},{"id":126,"by":127,"text":128},{"value":194,"source":93,"evidence":195},"Deepnote",[196],{"id":130,"by":131,"text":132},{"value":198,"source":93,"evidence":199},"Google Colab",[200],{"id":130,"by":131,"text":132},{"value":202,"source":93,"evidence":203},"PyCharm with Copilot",[204,208],{"id":205,"by":206,"text":207},"40858561","renewiltord","I just use PyCharm and Copilot plugin. Works like a charm.",{"id":209,"by":210,"text":211},"40860144","thehours","> Am I missing any? DataSpell: https:\u002F\u002Fwww.jetbrains.com\u002Fdataspell",{"comments":213,"answered":214,"key":215},26,12,[216,219,220],{"id":217,"text":218},"40858592","That's a cool blogpost! I'm mostly using Cursor now (just waiting until someone makes a kick-ass Emacs package so I can switch back!) so I can definitely see your perspective.…",{"id":59,"text":60},{"id":221,"text":222},"40858261","Thank you for the list - I think I've come across all of these in my research! I'll try highlight the differences for each. - https:\u002F\u002Fnoteable.io\u002F - as you say, it doesn't exist…",{"commenters":224,"max_depth":225,"maker_replies":213,"first_comment_mins":226,"lifespan_hours":227},32,6,27,46,[229,231,232,234,236,238,241,243,245,246,249,252,255,258,259,261,264,266,268,270,272,276,278,280,281,284,286,290,292,295,297,300,302,305,308,309,310,313,316,319,321,322,325,327,328,330,334,337,339,340,342,345,348,350,353,355,357,359,360,362,365,368,372,374,377,380,383,385,388,390,395,398],[130,34,34,230,131,34],"other\u002Fneutral\u002Fnone",[119,34,34,230,120,34],[233,34,34,230,180,34],"40858177",[235,34,34,230,120,34],"40858382",[217,34,34,237,11,41],"maker\u002Fneutral\u002Fnone",[239,34,34,240,120,34],"40866058","commenter\u002Fneutral\u002Fnone",[59,34,34,242,11,41],"maker\u002Fneutral\u002Fidea",[244,34,34,240,120,34],"40874191",[221,34,34,237,11,41],[247,34,32,248,131,34],"40858299","maker\u002Fpositive\u002Fdesign",[250,34,34,251,11,41],"40858806","maker\u002Fpositive\u002Fnone",[253,34,34,230,254,34],"40859225","spmurrayzzz",[256,34,34,230,257,34],"40859872","punkspider",[209,34,34,230,210,34],[96,41,32,260,97,34],"maker\u002Fpositive\u002Fexecution",[262,41,41,263,131,34],"40857897","project\u002Fcritical\u002Fnone",[265,41,34,237,11,41],"40858106",[267,41,34,240,131,34],"40858260",[269,41,34,237,11,41],"40858321",[271,41,34,240,131,34],"40858993",[273,41,45,274,275,34],"40860547","project\u002Fneutral\u002Fnone","lmeyerov",[277,41,34,237,11,41],"40860752",[279,41,34,237,11,41],"40858008",[156,39,45,242,157,34],[282,39,34,240,283,34],"40859434","westurner",[285,39,34,242,11,41],"40860998",[287,39,45,288,289,34],"40859632","project\u002Fneutral\u002Fidea","marcklingen",[291,39,34,237,11,41],"40860766",[293,39,45,288,294,34],"40865551","windexh8er",[103,45,32,296,104,34],"project\u002Fpositive\u002Fnone",[298,45,34,230,299,34],"40857625","skybrian",[301,45,34,237,11,41],"40857866",[303,45,34,230,304,34],"40864951","d4rkp4ttern",[306,45,34,307,11,41],"40857659","project\u002Fneutral\u002Fexecution",[137,35,45,242,138,34],[70,35,34,242,11,41],[311,32,45,274,312,34],"40857580","mritchie712",[314,32,34,230,315,34],"40857630","wolftickets",[317,32,34,230,318,34],"40857656","Bjartr",[320,32,34,237,11,41],"40857902",[163,225,34,230,164,34],[111,323,41,324,112,34],7,"project\u002Fcritical\u002Fidea",[326,323,34,240,149,34],"40862186",[141,18,45,242,142,34],[329,18,34,242,11,41],"40864149",[331,332,34,230,333,34],"40859324",9,"superkuh",[335,332,34,230,336,34],"40859935","vifon",[338,332,34,230,11,41],"40860887",[205,33,34,230,206,34],[341,33,34,237,11,41],"40858677",[343,33,32,344,206,34],"40858838","project\u002Fpositive\u002Fdesign",[346,33,34,347,11,41],"40859521","maker\u002Fneutral\u002Fdesign",[126,349,34,230,127,34],11,[351,349,34,352,149,34],"40862192","other\u002Fcritical\u002Fexecution",[354,349,34,230,127,34],"40862452",[356,349,34,230,149,34],"40870839",[358,349,34,237,11,41],"40864183",[115,214,41,324,116,34],[361,214,34,237,11,41],"40860812",[179,363,45,364,180,34],13,"project\u002Fneutral\u002Fperformance",[366,363,34,367,11,41],"40858798","project\u002Fpositive\u002Fperformance",[369,370,34,230,371,34],"40858929",14,"mathiasn",[373,370,34,237,11,41],"40860836",[148,375,41,376,149,34],15,"project\u002Fcritical\u002Fexecution",[378,375,34,379,11,41],"40864237","maker\u002Fneutral\u002Fexecution",[172,381,45,382,173,34],16,"project\u002Fneutral\u002Fdesign",[384,381,34,382,11,41],"40864160",[386,387,45,274,299,34],"40857587",17,[389,387,34,274,11,41],"40857611",[391,392,34,393,394,34],"40860421",18,"other\u002Fcritical\u002Fidea","localfirst",[396,392,34,397,11,41],"40860917","project\u002Fpositive\u002Fidea",[399,392,32,397,149,34],"40872013",[],[402,407,412,417,421,425],{"id":403,"name":404,"points":405,"category":406},"40843867","Pikimov",1150,"creative",{"id":408,"name":409,"points":410,"category":411},"40837610","SmoothMQ",656,"infrastructure",{"id":413,"name":414,"points":415,"category":416},"40832146","Docmost",551,"productivity",{"id":418,"name":419,"points":420,"category":20},"40869877","Xcapture-BPF",447,{"id":422,"name":423,"points":424,"category":20},"40847699","Doggo",400,{"id":426,"name":427,"points":428,"category":416},"40846242","Edna",331,false,"v2",{"id":432,"name":5,"line":6,"ai":429,"cat":20,"year":13,"points":14,"comments":15,"rec":29,"s":433,"top":18,"kw":434,"maker":11,"domain":10,"tier":19,"attempts":34,"strip":435,"url":9,"mix":436,"praise":92,"objection":108,"sub":21},40857009,0.01,[24,25,26,27,28],"0000050030030330500330013000500013013305",[32,33,34,35],1794036657139]