Apple Core AI (.aimodel) conversions of the Laya System 1 decision models for macOS 27 on Apple Silicon, with measured drift against Laya's own runtim
Ahmed Alaaeldin
ahmedandaloes
AI & ML interests
None yet
Recent Activity
updated a dataset 2 days ago
ahmedandaloes/security-decisions published a dataset 2 days ago
ahmedandaloes/security-decisions updated a collection 2 days ago
Laya — Core AI (Apple)Organizations
None yet
MLX (Apple Silicon) conversions of the Decision 3.0 System One decision models — runnable with mlx-vlm.
MLX 4/6/8-bit + bf16 of ZySec-AI/SecurityLLM for Apple Silicon.
MLX 4/6/8-bit + bf16 of fableforge-ai/NEXUS-Security for Apple Silicon. GGUF: King3Djbl/nexus-security-GGUF.
Apple Core AI (.aimodel) conversions of the Decision 3.0 family for macOS 27 and the Neural Engine.
MLX 4/6/8-bit of empero-ai/Qwythos-9B-v2 (qwen3_5) for Apple Silicon. GGUF: empero-ai/Qwythos-9B-v2-GGUF.
MLX 4/6/8-bit + bf16 of viettelsecurity-ai/security-llama3.2-3b for Apple Silicon.
-
ahmedandaloes/security-llama3.2-3b-MLX-4bit
Text Generation • 3B • Updated • 48 -
ahmedandaloes/security-llama3.2-3b-MLX-6bit
Text Generation • 3B • Updated • 46 -
ahmedandaloes/security-llama3.2-3b-MLX-8bit
Text Generation • 3B • Updated • 83 -
ahmedandaloes/security-llama3.2-3b-MLX-bf16
Text Generation • 3B • Updated • 42
MLX conversions (4/6/8-bit + bf16) of oyildirim/CyberStrike-OffSec-35B for Apple Silicon.
-
ahmedandaloes/CyberStrike-OffSec-35B-MLX-4bit
Text Generation • 35B • Updated • 168 • 1 -
ahmedandaloes/CyberStrike-OffSec-35B-MLX-6bit
Text Generation • 35B • Updated • 125 • 1 -
ahmedandaloes/CyberStrike-OffSec-35B-MLX-8bit
Text Generation • 35B • Updated • 63 • 1 -
ahmedandaloes/CyberStrike-OffSec-35B-MLX-bf16
Text Generation • 35B • Updated • 63 • 1
Apple Core AI (.aimodel) conversions of the Laya System 1 decision models for macOS 27 on Apple Silicon, with measured drift against Laya's own runtim
Apple Core AI (.aimodel) conversions of the Decision 3.0 family for macOS 27 and the Neural Engine.
MLX (Apple Silicon) conversions of the Decision 3.0 System One decision models — runnable with mlx-vlm.
MLX 4/6/8-bit of empero-ai/Qwythos-9B-v2 (qwen3_5) for Apple Silicon. GGUF: empero-ai/Qwythos-9B-v2-GGUF.
MLX 4/6/8-bit + bf16 of ZySec-AI/SecurityLLM for Apple Silicon.
MLX 4/6/8-bit + bf16 of viettelsecurity-ai/security-llama3.2-3b for Apple Silicon.
-
ahmedandaloes/security-llama3.2-3b-MLX-4bit
Text Generation • 3B • Updated • 48 -
ahmedandaloes/security-llama3.2-3b-MLX-6bit
Text Generation • 3B • Updated • 46 -
ahmedandaloes/security-llama3.2-3b-MLX-8bit
Text Generation • 3B • Updated • 83 -
ahmedandaloes/security-llama3.2-3b-MLX-bf16
Text Generation • 3B • Updated • 42
MLX 4/6/8-bit + bf16 of fableforge-ai/NEXUS-Security for Apple Silicon. GGUF: King3Djbl/nexus-security-GGUF.
MLX conversions (4/6/8-bit + bf16) of oyildirim/CyberStrike-OffSec-35B for Apple Silicon.
-
ahmedandaloes/CyberStrike-OffSec-35B-MLX-4bit
Text Generation • 35B • Updated • 168 • 1 -
ahmedandaloes/CyberStrike-OffSec-35B-MLX-6bit
Text Generation • 35B • Updated • 125 • 1 -
ahmedandaloes/CyberStrike-OffSec-35B-MLX-8bit
Text Generation • 35B • Updated • 63 • 1 -
ahmedandaloes/CyberStrike-OffSec-35B-MLX-bf16
Text Generation • 35B • Updated • 63 • 1