High-quality datasets targeting reinforcement learning, SFT, cybersecurity theory, and code-as-action systems.
👋 Open to Work
True Debreuil
True2456
AI & ML interests
Optimising LLM's for the Mac Ecosystem & Unmanned Vehicles. AUS
Recent Activity
new activity about 2 hours ago
True2456/gemma-4-12b-it-qat-4bit-frontierdistill-v2:Excellent speed liked a dataset 3 days ago
tate01684/Vibe-Coding-Claude-Fable-5Organizations
Large Mixture-of-Experts (MoE) and pruned models quantized and optimized for Apple Silicon via the MLX framework.
-
True2456/DeepSeek-V4-Flash-0731-REAP37-native-MLX
Text Generation • 31B • Updated • 1.27k • 8 -
True2456/Mati-3.7-173B-4.6bit-MLX
Image-Text-to-Text • 29B • Updated • 66 -
True2456/nemotron-3-super-120b-cybersecurity-theory-lora-mlx
Text Generation • Updated • 2 -
True2456/nemotron-3-super-120b-a12b-theory-lora-mlx
Text Generation • Updated
High-quality datasets targeting reinforcement learning, SFT, cybersecurity theory, and code-as-action systems.
Large Mixture-of-Experts (MoE) and pruned models quantized and optimized for Apple Silicon via the MLX framework.
-
True2456/DeepSeek-V4-Flash-0731-REAP37-native-MLX
Text Generation • 31B • Updated • 1.27k • 8 -
True2456/Mati-3.7-173B-4.6bit-MLX
Image-Text-to-Text • 29B • Updated • 66 -
True2456/nemotron-3-super-120b-cybersecurity-theory-lora-mlx
Text Generation • Updated • 2 -
True2456/nemotron-3-super-120b-a12b-theory-lora-mlx
Text Generation • Updated
Curated collection of Gemma-4 MLX optimizations, on-policy alignment corrections, and agentic fine-tunes.