https://modu-print.com/tags/omo/Hugoko-KRSat, 28 Feb 2026 01:58:29 +0900https://modu-print.com/posts/16150/Sat, 28 Feb 2026 01:58:26 +0900https://modu-print.com/posts/16150/맥 스튜디오 M3 Ultra 512g 모델로 이런짓 저런짓을 좀 해보고 있는 와중에 중간 중간 관련 글을 정리해 올려보고 있습니다. Bash pirogom@pirogom-Mac-Studio ~ % lms ls You have 12 models, taking up 599.66 GB of disk space. LLM PARAMS ARCH SIZE DEVICE google/gemma-3-27b (1 variant) 27B gemma3 16.87 GB Local llama-3.3-70b-instruct@4bit 70B Llama 39.71 GB Local llama-3.3-70b-instruct@8bit 70B Llama 74.98 GB Local llama-4-scout-17b-16e@4bit 17B llama4 61.14 GB Local llama-4-scout-17b-16e@8bit 17B llama4 115.46 GB Local minimax/minimax-m2.5 (1 variant) 230B minimax_m2 128.68 GB Local qwen2.5-coder-3b-instruct-mlx 3B Qwen2 1.75 GB Local qwen3-coder-next@4bit qwen3_next 44.86 GB Local qwen3-coder-next@8bit qwen3_next 84.67 GB Local zai-org/glm-4.6v-flash (1 variant) 9B glm4v 7.09 GB Local zai-org/glm-4.7-flash (1 variant) 30B glm4_moe_lite 24.36 GB Local EMBEDDING PARAMS ARCH SIZE DEVICE text-embedding-nomic-embed-text-v1.5 Nomic BERT 84.11 MB Local pirogom@pirogom-Mac-Studio ~ % lms ls You have 12 models, taking up 599.66 GB of disk space. LLM PARAMS ARCH SIZE DEVICE google/gemma-3-27b (1 variant) 27B gemma3 16.87 GB Local llama-3.3-70b-instruct@4bit 70B Llama 39.71 GB Local llama-3.3-70b-instruct@8bit 70B Llama 74.98 GB Local llama-4-scout-17b-16e@4bit 17B llama4 61.14 GB Local llama-4-scout-17b-16e@8bit 17B llama4 115.46 GB Local minimax/minimax-m2.5 (1 variant) 230B minimax_m2 128.68 GB Local qwen2.5-coder-3b-instruct-mlx 3B Qwen2 1.75 GB Local qwen3-coder-next@4bit qwen3_next 44.86 GB Local qwen3-coder-next@8bit qwen3_next 84.67 GB Local zai-org/glm-4.6v-flash (1 variant) 9B glm4v 7.09 GB Local zai-org/glm-4.7-flash (1 variant) 30B glm4_moe_lite 24.36 GB Local EMBEDDING PARAMS ARCH SIZE DEVICE text-embedding-nomic-embed-text-v1.5 Nomic BERT 84.11 MB Local