MSVC+CUDA llama.cpp fork for Windows: KDA/GDN, long-context KV placement, TurboQuant/TCQ when measured, speculative draft. Lab defaults — TriAttention opt-in only, not recommended.
-
Updated
Jul 28, 2026 - C++
MSVC+CUDA llama.cpp fork for Windows: KDA/GDN, long-context KV placement, TurboQuant/TCQ when measured, speculative draft. Lab defaults — TriAttention opt-in only, not recommended.
Allows you to do regression on game journals to determine effective loadouts.
Pact Language Support for JetBrains IDEs 🚀
Python package for the Kemeny Decomposition Algorithm (KDA) together with some Markov chain tooling.
Github action for polling status with timeout
Kadena Wallet - self-custody desktop wallet for KDA. Native Windows / macOS / Linux. MIT-licensed.
Evidence-first optimization of numerically exact KDA training on NVIDIA GB10
Accelerate Kimi Delta Attention computations with high-performance CUTLASS kernels designed for NVIDIA SM90 architectures and beyond.
To associate your repository with the kda topic, visit your repo's landing page and select "manage topics."