🔄 In a Training Loop

Pasha

pashak

prism-ml

·

AI & ML interests

Probabilities

Recent Activity

new activity 1 day ago

prism-ml/Ternary-Bonsai-8B-gguf:Q2_0 g64 ternary is ~5× slower than Q4_K_M on Samsung S22 CPU - missing optimized kernel?

new activity 14 days ago

prism-ml/Ternary-Bonsai-27B-gguf:Benchmark: 27B Q2_0 vs Qwen3.6-27B IQ2_XXS on one RTX 5060 Ti 16 GB (quality@budget, speed, VRAM, energy)

new activity 16 days ago

prism-ml/Ternary-Bonsai-27B-gguf:docs: point the llama.cpp quickstart at the PQ2_0 pack

View all activity

Organizations

pashak 's models

None public yet