Mac users: if you are running Qwen models locally, make sure you use MLX (Apple's version of PyTorch, sort of). It's the only way to get decent performance.
Windows users: just buy a GeForce RTX ffs
stevibe@stevibeUpdate: Qwen3.5:9b head-to-head (MLX, Apple Silicon optimized) Mac Studio M2 Ultra: 89.74 tok/s Mac Mini M4: 20.82 tok/s MLX basically doubles the speed on both machines.
Sending doesn't need an account. Sign in if you want a private collection.
Preview another link
Works with X, Instagram, TikTok & YouTube.
Sending doesn't need an account. Sign in if you want a private collection.
Preview another link
Works with X, Instagram, TikTok & YouTube.
Due to the leakage of Claude Code, the following projects were generated: https://github.com/instructkr/claw-code, Star 94.3K, Fork 88.5k https://github.com/claude-code-best/claude-code, Star 5.7K, Fork 7.2k https://github.com/paoloanzn/fr…
Anne Widdecombe was killed with a hammer. Rhiannon Whyte with a screwdriver. Wayne Broadhurst with a knife. Henry Nowak with a ceremonial blade. Jason Arday with the truth.
@shiroko_bridge それはそう ただ、初心者の体験としては最高に楽しかった思い出がある！ 何も考えずに ollama run gemma3:12b だけで初めて自分のpcが喋ったのは楽しかった！！！！！！！
日本のlocal llmの投稿を見るとよくollamaを使っているのがみられますが、やめましょう。 ollamaを使用していながら速度が遅いというのは、軽自動車で130km/hでないといってるのと同じです。 linux+llama.cppに乗り換えましょう。 数十%速度上がるよ
This might be the most beautiful thing I've read in my life. What a privilege to be alive in 2026. https://ordinaryabundance.com
The more nonsensical and obviously absurd a belief is, the more people who hold it (or pretend to) will resist recanting. The potential for humiliation rises in direct proportion to the ridiculousness of what they've publicly swallowed. Th…
