Avatar
😀

Organizations

  • Ini adalah list yg aku buat based on apa aku dpt dari community lain dan list apa local community kita buat, local community focus more on llm dan lambat untuk explore sendiri benda baru Sebelum ni apa aku buat based on apa community lain buat vectordb 14/3/23 - https://twitter.com/khursani8/status/1635490473312919553 voice conversion 6/2/23 - https://x.com/khursani8/status/1622548019706216448?s=46&t=KgPtmqLaM0tK9HeoMI1fZg stable diffusion 1/3/23 - https://twitter.com/khursani8/status/1630776772907388930 llama qlora finetune 21/7/23 - https://twitter.com/khursani8/status/1682397593756917761 clean llm data 25/7/23 - https://twitter.
    Created Fri, 10 May 2024 07:54:52 +0800
  • Introduction Semalam Facebook release llama3, macam biasa dia hanya fokus dekat English dan juga English benchmark. Mula2 nak try buat merging tapi tak ada model lagi yang dah finetune llama3 untuk Bahasa Melayu. So aku finetune sendiri dan apply instruct vector aku extract dari llama3 instruct supaya model base llama3 yang dah finetune ke Malay boleh follow instruction dan jawab dalam Malay. Apa beza Base, Instruct dengan Chat? Dekat atas ni contoh flow untuk dapatkan Chat model
    Created Sat, 20 Apr 2024 07:10:45 +0800
  • Buat 3d model pakai nerfstudio Pengenalan Seminggu sebelum CNY aku kena paksa gi melancong, sambil melancong terjumpa benda menarik so aku just collect dataset sambil berjalan Aku dari dulu nak buat 3d model pakai nerf atau gaussian splatting cuma tak keluar rumah, so sambil keluar rumah boleh laa scrape real world Apa tu 3D Reconstruction? 3D reconstruction adalah process untuk create 3d representation objek atau scene daripada gambar 2d atau sensor.
    Created Sat, 17 Feb 2024 12:23:36 +0800
  • Pakai Sadtalker utk lipsync Pembuka Hari ini AI untuk rakyat telah diluncurkan, benda pertama yang menarik perhatian aku(sebab tak dapat register website dia lol) adalah AI announcer? A.K.A Intan dimana dia combine Text to Speech(TTS) dan syntesis gambar. So far based on experience aku, benda ni probably pakai SadTalker atau adik beradik dia. Apa tu SadTalker? SadTalker adalah model untuk buat animation terhadap gambar kepala yang menggunakan audio suara sebagai input.
    Created Tue, 16 Jan 2024 18:39:40 +0800
  • 1.https://cs.stanford.edu/~zxie/textgen.pdf In Introduction the author talk about the tradeoff between Rule-based and Neural e2e method ![null](/img/Screen Shot 2018-10-24 at 7.43.01 PM.png) WIP
    Created Wed, 24 Oct 2018 11:40:17 +0000
  • https://www.kaggle.com/sinakhorami/titanic-best-working-classifier?scriptVersionId=560373 https://www.kaggle.com/nadare/titanic-randomforest-0-79425 winning solution https://www.kaggle.com/sudalairajkumar/winning-solutions-of-kaggle-competitions
    Created Mon, 02 Jul 2018 14:35:26 +0000
  • https://medium.com/@alark/we-need-more-interactive-data-visualization-tools-for-the-web-in-python-ad80ec3f440e https://www.data-to-viz.com/caveats.html https://medium.com/@matthijsmali/data-visualization-with-the-elk-stack-69d583b636b9 http://learnjsdata.com/
    Created Thu, 14 Jun 2018 15:41:22 +0000
  • At first I install kaggle-api so that I can download the datasets more easily. I decide to participate in this competition first since it was the first one being teach in fast.ai class. After download the datasets I check at the file and found out that there is two folder which is train and test folder. Inside the folder got images of cats and dogs in one folder. It was better to seperate all those image in different folder and use that folder name as label.
    Created Sat, 02 Jun 2018 06:07:50 +0000
  • TODO http://anotherdatum.com/who-wants-to-be-a-coder.html DONE
    Created Fri, 11 May 2018 15:01:28 +0000
  • lambda cannot upload s3 https://stackoverflow.com/questions/35589641/aws-lambda-function-getting-access-denied-when-getobject-from-s3 improving node performance https://www.smashingmagazine.com/2012/11/writing-fast-memory-efficient-javascript/
    Created Tue, 06 Mar 2018 07:14:50 +0000
Next