Kamarul Adha

Exciting updates! Malaysia AI just published preprint on our work for safe for work classifier, check it out here arxiv.org/abs/2407.20729

@belleyangcerdik I win.

Call me this!

@belleyangcerdik Oh my goodness you really going to bring that up here? Now? Sorry la sidai. Multiple meetings/presentations to handle this week.

@kamaaarul Selagi sidai selagi tu xmau 😏😏😏

Backend hancur Frontend cun

@belleyangcerdik Ahhhh no wonder cun. Maaf, saya HTMX enjoyer sahaja. Teach me akak belle

@kamaaarul Sebab guna vanilla JavaScript 😏🔥

@oscarnazhan @hlmshkr Marilah kita sama2 setup your CUDA needs on windows.

me in a nutshell meme by @kamaaarul @hlmshkr

Data Engineering 101 Kita berpendapat, ada gap dalam market dalam skills data engineering. Jadinya, kita akan jalankan kelas 8 minggu. Kita telah bincang dgn Azwan, Kamarul, Riza, Dr Rahman dan Dr Syarmine sebagai trainer. Syllabus ada dlm GDrive: drive.google.com/file/d/1-CO_Da… (1)

Of the many AI projects done in Malaysia, these guys I can vouch are as legit as they come. Hopefully this will be a one off by our minister. We have amazing local talents. You need to empower them.

MESOLITICA @ MINISTRY OF DIGITAL We had a session with YB Gobind yesterday morning together with the agencies under Ministry of Digital. A special briefing session sharing the state of AI in 2024, current R&D trends in LLM and sharing our work thus far including upcoming release of MaLLaM 🌙 2.0 and Malaysia AI (MYAI) volunteer team work on “SFW Classifier” for AI Safety. Together with CTO Husein, 6 of our core contributors from MYAI was in attendance: Aisyah Razak, Kamarul Adha, Ariff Nazhan, Halim Shukor, Faiq & Mas Aisyah. 🤖🧠🌊 ps: After streamlining the “wish list” at the national level, we are ready to execute if we are tasked to do so with related statekholders towards AI Sovereignty for Malaysia.

Bismillahhh, May Allah ease your way MalaysiaAI 💪🏻

this is it! way to go meso and mlaysia ai team!

We are so back

Perbincangan yang menarik bersama delegasi dari Mesolitica, sebuah pembangun Kecerdasan Buatan Multimodal bagi Asia Tenggara. Kunjungan yang diketuai oleh Tn. Khalil Nooh, Pengasas Bersama dan CEO, serta Tn. Husein Zolkipli, Ketua Pegawai Teknologi membincangkan pelbagai aspek Generative AI, Model Bahasa Raya (Large Language Model - LLM) dan model yang dibangunkan oleh Mesolitica, MaLLaM 2.0. Malaysia Large Language Model ini merupakan model bahasa pintar (chat language model) yang dibangunkan oleh bakat-bakat tempatan, dan berupaya melaksanakan tugas-tugas pemprosesan bahasa semulajadi, seperti menjawab soalan dan berkomunikasi menggunakan bahasa-bahasa di Malaysia termasuk Bahasa Melayu, Tamil, Mandarin dan Jawi. Dilatih menggunakan set-set data tempatan, MaLLam 2.0 dibangunkan dengan tujuan mencontohi cara rakyat Malaysia bercakap dan berkomunikasi. Saya kagum dengan usaha Mesolitica untuk menjadi antara pemacu LLM yang unggul di Asia Tenggara dan pihak Kementerian bersedia memberi sokongan kepada mana-mana inisiatif yang selaras dengan agenda digital negara. --- A most interesting meeting with the team from Mesolitica, a Multimodality Artificial Intelligence developer for Southeast Asia. Led by Mr. Khalil Nooh, Co-founder and CEO; and Mr. Husein Zolkipli, Chief Technology Officer, we were given an insight into the world of Generative AI, Large Language Models (LLM) and Mesolitica’s homegrown AI model, MaLLaM 2.0. MaLLaM 2.0 is a multilingual Malaysian chat language model developed by local talents. It is capable of performing natural language processing (NLP) tasks, such as answering questions and engaging in conversations in local languages, including Malay, Tamil, and Mandarin. Trained on Malaysian datasets, MaLLaM 2.0 is designed to better mirror the language and communication style of an everyday Malaysian. I am impressed by Mesolitica’s initiatives as an important LLM player in the region, and the Ministry of Digital is happy to support initiatives that are aligned to the digital empowerment of our country. #MalaysiaMADANI #ICT #MaLLaM #LLM

MaLLaM NPM package now supports generating embeddings from text! npmjs.com/package/mallam…

@SakiHanajima__ Aamiin. I will try. 🌹

kamarul boleh jadi lecturer hahaha

Join me and my team (MYAI) tonight at Devtalk where we will explore RAG using Mallam. Tonight's session will be more hands-on and beginner friendly.

These two crazy mentally unstable kittens really be testing my patience. Kalau tak menyakat, memang tak sah diorang ni. @aisyahhhrzk @oscarnazhan

Interested to learn more about RAG (Retrieval-Augmented Generation)? @kamaaarul will share more about RAG that is being used with MaLLaM by @mesolitica, tomorrow (23 Jun, Sunday), at 9.30PM, at DevTalk Discord server!

Explore RAG dengan @kamaaarul esok!

@belleyangcerdik @nekoketew Sorry la awok. Mase tu busy with work. This weekend pon dah fully booked. Nanti I find a suitable date and we go eat together ramai2 okay. 🌹

@kamaaarul @nekoketew Xnak berharap dh dgn kamarul. Asyik kena sidai je. Buat plan nnt mesti kena sidai. Janjimu tidak bararti bagiku.