GitHub · Python
synchopate/llama-modchip
Zero-GPU CPU inference for 30B-284B MoE models (Qwen3.8 176.5B cold-cache certified in 28.26/48.35 GiB Cgroup v2, DeepSeek-V4 284B, Nemotron-30B) + RouteCrack 2.0.
- Stars
- 1,402
- Forks
- 138
- Stars gained in 24 hours
- +19
- Forks gained in 24 hours
- +3
Topics: ai-agents, apple-silicon, avx512, claude-code, cpu-inference, deepseek, deepseek-v4, gguf, hpc, llama-cpp
Latest data point: Nov 22, 2026 · 1,402 stars · 138 forks