Fatty Singapore AI Base Model – Beta Trial Testing Version 1.6 – A Micro Local SG Artificial intelligence
Fatty Singapore AI Base Model ( Artificial intelligence ) – Beta Trial Testing Version 1.6 – Supervised Machine Learning Model – Users may use the model for inference only. No fine-tuning, distillation, or derivative works allowed without written permission.
A 42M Parameter AI Singapore Soul in the Making Local Clean From Scratch – Build by Singaporean
The Fatty AI Method: Pure Supervised Learning
Fatty AI is built from the ground up using Supervised Machine Learning. This is the core of how it thinks, but it’s a very different process from what you might have heard from large AI companies.
Fatty will only expand tools after the base is stable with more learning.
What is Supervised Learning?
In simple terms, Supervised Learning is like teaching a child with flashcards.
You show the child a flashcard with a Question (the input).
On the back is the correct Answer (the desired output).
By studying thousands of these flashcards, the child learns the underlying patterns and eventually can answer new, similar questions on their own.
This is exactly how Fatty AI was trained + UI testing and check model response
How Fatty AI Was Supervised?
I acted as the teacher. For over 600,000 examples, approximates around 786 million characters in this version which is still underfed. I provided the model with a perfect pair:-
This Model is heavily trained on routes and address and underfed on most daily use conversations on this Trial Testing Version. Estimated growth with minimum 5 million training examples just for local in 2026.
It does help tourist to navigate for around 19 to 20 Hotels at Orchard Road, Tanglin Rd, Scotts Rd to Malls right now. Examples: How to go Tudor Court from Hyatt Hotel? How to go Yotel from Plaza Singapura? How to go Paragon? etc.
A precise instruction or question ( e.g., How do I walk from Concorde Hotel to Camden Medical? )
The exact, correct answer I wanted it to learn (e.g., Starting point: Concorde Hotel… [ – detailed route – ]…)
The model’s only job was to analyze these millions of Question and Answer, pairs and discover the deep patterns that connect them. Through this intense, focused training over hundreds of epochs, it didn’t just memorize, it learned to understand the structure of navigation questions, the properties of locations, and how to generate a correct response.
The Key Difference: Deep Education vs Surface Level Manners
This is where Fatty AI differs completely from many models.
Fatty AI – Deep Education
Step 1: Foundation Learns from my clean, hand written prompts. Its entire worldview is built from high quality, structured knowledge.
Step 2: Supervised Phase, This is its entire life. Supervision is the primary, continuous process that builds its core intelligence.
Result:- A specialist with a pure and reliable knowledge base, capable of true reasoning within its domain.
Like many solo developers, whether fine tuning a model or building a base model from absolute scratch, the early days were pure torture. When my training data was only 2,000 to 3,000 lines, every epoch felt pointless. I wanted to give up a thousand times. But giving up was never an option. So I kept pushing, day after day, night after night, since July 2025, Fatty’s actual birthday month. Yes, you read that right: this is a genuine solo built base model, hand fed by one stubborn Singaporean who refused to quit.
My personal point of view on building a model like Fatty? It is affordable but very time consuming behind the scenes. It is not about how well one can write programming languages, C++, Pascal, Java or whatever, to stitch responses or pull data. What is this about? It’s not a tool right now, tools can be added easily in future.
What am I looking for? To complete something the first batch of scientists decades ago could not finish because they lacked the right hardware or tools. I want to make their wish work too.
This micro or mini base model is not trained with intents, tags, definitions, or inserted placeholders for stitching and pulling data. It doesn’t make sense anymore because it is not a scraping machine, that would derail my thoughts on continuity.
What is Fatty AI?
Fatty AI is an ambitious experiment: a compact base model built entirely from scratch. No pre-training, no scraping, No tags, no intent training, no placeholders. Every piece of knowledge was carefully hand fed, not web scraped, ensuring zero copyright risks. It’s designed to run locally on Android phones ( FP32, from 4GB to 16gb+6gb RAM ), with the goal of becoming a true Singapore expert in your pocket.
A Micro or Mini Base, simple and yet consist of something what Singaporean remember. – Give me more suggestions such as ‘Mata last time wear Shorts Lar’ to this email. With your help, the model can further learning with more different brains of Singapore.
Our Approach: A Different Path
The common belief is that advanced AI requires immense resources. We’re exploring a different path. By prioritizing relentless focus and over 600,000 hand crafted lessons over financial capital, Fatty AI is our testbed for a more efficient AI paradigm.
While conventional wisdom relies on massive data centers, we chose depth over breadth. Instead of scraping the entire internet, we wrote those 600,000+ precise lessons estimated around 786 million characters. The result is a compact model that aims to prove focused, efficient, and capable AI is a viable path forward.
Current Knowledge Coverage (Beta Trial Testing Version 1.6)
Please note, this is a beta version for trial testing and has limitations and it is underfed especially for chit-chats. Coverage is a work in progress.
Core Infrastructure: ~90% address coverage of government schools, police posts ( NPP – Neigbourhood Police Post and NPC – Neigbourhood Police Centre ), hotels, and shopping malls.
Local Food & Culture: ~50% coverage (and growing), including dishes like chicken rice, laksa, roti john, chwee kueh, curry, roti prata etc.
Landmarks: including Marina Bay Sands, Resort World Sentosa, Sentosa , Farms etc.
Orchard Road: ~47% route coverage in this beta. 100% completion for Orchard Road is our immediate focus. (This version Orchard road route are around 20 hotels to 90% of the shopping centre / mall covering from Tanglin Rd to Orchard Rd)
This is a Learning System, Not a Finished Product. However, it still can assist tourist on many Orchard routes from Hotel to Shopping Centre or Shopping Mall.
This is a different kind of AI project in Singapore. Because it’s taught with clean, structured data, we’ve observed promising signs of early reasoning abilities, like adapting its knowledge to new questions. However, Beta 1.6 is underfed and can sometimes produce incorrect or unexpected responses. It doesn’t just recall it’s learning to understand, and it needs your help to improve.
Join Us on This Journey
If our work resonates with you and you feel moved to support our path, we would be deeply grateful. Your support, at any level, is a sign of shared belief and keeps us going.
Fuel Our Journey with a Coffee
A Gesture of Solidarity – Typically from $1 to $10
Your contribution, like friends, is a warm gesture that tells us we’re on the right path. It means the world to us. – This is a long and arduous path. Your support directly fuels this foundational work.
Help Us Build a Milestone
A Meaningful Contribution – Starting from $100
This level of support allows us to dedicate significant time and resources to key features.
Become a Cornerstone of Our Vision
Typically starting from $1,000 to support us for continuing this project which is a very long path ahead as you may know this is a fully hand crafted base model with no pretrained, no wrapping and without online scraping.
Beta Trial Testing Version 1.6 is a stepping stone. Your support is not for servers, but for the dedicated time needed to push this independent project forward to complete Orchard Road and lay the groundwork for a more stable Beta 2.5 in 2026.
Support the Development of Fatty AI
Fatty AI is a project from Yi Pin Qian, built by a single developer. It’s a passion project that questions whether AI must always be big and corporate to be smart.
Your support allows for the full time dedication required to solidify its knowledge base and improve its reasoning. Supporting Fatty AI is a vote of confidence for an alternative, more focused path in AI.
How to Support (Our Suggestions)
We believe in genuine resonance and offer the following avenues, but you have complete freedom.
We fully understand and respect the spirit of free giving. Every contribution, regardless of size, is a direct investment into the core research and development of Fatty Singapore AI.
Support Channel:
PayPal: paypal.me/supportfatty
PayNow UEN: 53326052C
For Strategic Partnerships & Investment:-
Fatty Singapore AI is more than a passion project, it is a viable, proven blueprint for building efficient and deeply capable AI. While community support fuels our day to day progress, we are now seeking strategic partners and investors who share our long term vision to scale this paradigm.
We believe the future of AI is not just bigger, but smarter and more focused. Our methodology – Deep Feed Learning with high quality, hand crafted data that has been validated. The next step is to apply this blueprint to new domains and scale our technology.
We are open to discussions regarding:
Strategic Investment to accelerate R&D and expand into new verticals.
Commercial Licensing of our technology and data construction methodology.
Investors are welcome to send an email to fatty@yipinqian.com with your phone number to contact, thank you for your understanding.
Research Collaboration with institutions exploring the frontiers of efficient model training.
If you see the potential in a more transparent, focused, and human centric path for Fatty Singapore AI, we would be honored to speak with you.
Let’s build the future of intelligent systems, together. (All Suggestion welcome)
Contact us at: fatty@yipinqian.com
Fatty’s Growth Updates Click Here
Fatty’s account page on Github -> https://github.com/YiPinQian/Fatty
Download Here: Download Link For Trial Testing Version 1.6 – Fatty
Fatty’s home: https://fatty.sg + https://fatty.com.sg + https://yipinqian.com/fatty
© 2025 Yi Pin Qian – All rights reserved – Model weights licensed for non-commercial inference only. No reproduction, modification, or distribution without consent. – The model contains permanent identifiers that will reveal its true origin in any unauthorised copy.
Fatty’s email: fatty@yipinqian.com
JA 
EN 
ZH 
MS 
ID 
TA 
HI 
VI