06:49
Nanowhale-100m: Fascinating Implemention of DeepSeek-V4 Architecture
Fahd Mirza walks through Nanowhale-100M, a 110 million parameter language model built entirely from scratch—no borrowed weights—that...
Sort by: Newest Items
There are 129 items in this page
10:33
Mellum2: JetBrains’ New Coding Model – vLLM + MCP Tool Use Locally
JetBrains has released Mellum 2, a 12-billion-parameter mixture-of-experts coding model that runs at the compute cost of a 2.5-billio...
14:00
Master Ideogram 4: 5 Reasons Why Ideogram 4 is the New King of Graphic Design
This tutorial from the Veteran AI channel provides a detailed technical walkthrough of Ideogram 4, the first open-weight text-to-imag...
13:43
Gemma4 12B in Quantization-Aware Training (QAT) with Ollama – Full Testing
Google's Gemma 4 12B model now has Quantization-Aware Training (QAT) checkpoints, and Fahd Mirza puts them through a full workout in...
12:30
Ideogram 4: World’s Best Text-to-Image Model? Let’s Test Locally
Fahd Mirza installs and tests Ideogram 4 locally, providing a candid assessment of its real-world hardware requirements and architect...
12:27
Nemotron 3 Ultra – NVIDIA’s Most Powerful Open Model – Long Running Agents
NVIDIA has released Nemotron Ultra, its largest open model to date at 550 billion total parameters—with only 55 billion active at inf...
11:51
Nvidia Just Introduced 4 New Stunning AI Updates
At Nvidia GTC Taipei, CEO Jensen Huang unveiled four major updates that collectively push Nvidia deeper into open-source AI, agentic...
13:02
MiniMax M3: Frontier Coding, 1M Context, Native Multimodality – Thorough Testing
Fahd Mirza puts MiniMax M3 through a hands-on evaluation, opening with a striking demonstration: a single prompt produces a fully sel...
08:12
LocateAnything: NVIDIA’s New AI Sees EVERYTHING: Run Locally
NVIDIA's LocateAnything is a 3-billion-parameter vision-language model that goes beyond standard image classification to precisely lo...
11:06
Best Qwen3.6 Quant You Can Run Right Now Locally
Fahd Mirza examines Nvidia's official FP4 quantization of Qwen3 35B A22B — a release validated by Nvidia's own model optimizer tool a...