Pinned Loading

1. atomic-llama-cpp-turboquant atomic-llama-cpp-turboquant Public

   Forked from AtomicBot-ai/atomic-llama-cpp-turboquant

   llama.cpp fork with TurboQuant WHT-rotated KV cache & weight compression + Gemma 4 MTP and Qwen 3.6 NextN speculative decoding (+30-50% throughput).

   C++ 1

2. llama.cpp-turboquant-hip-tuned llama.cpp-turboquant-hip-tuned Public

   llama.cpp-turboquant-hip-tuned

   C++

3. codebuff-swe-bench codebuff-swe-bench Public

   Forked from CodebuffAI/codebuff-swe-bench

   Harness used to benchmark Codebuff against SWE Bench benchmarks

   Python

4. metacognitive-architect metacognitive-architect Public

   Forked from mihazs/metacognitive-architect

   An Advanced Prompt Architecture for Stateful AI Agents

5. mtp-baker mtp-baker Public

   Python