3 ms·
ollama runs deepseek-r1:7b on AMD 8945HS, CPU-only, at ~12 tokens/s. You can get started pretty easily in the ~7B model range, for learning purposes.
by yencabulator 2y ago
ollama runs deepseek-r1:7b on AMD 8945HS, CPU-only, at ~12 tokens/s. You can get started pretty easily in the ~7B model range, for learning purposes.