RightNow
Popular repositories Loading
-
autokernel
autokernel PublicAutoresearch for GPU kernels. Give it any PyTorch model, go to sleep, wake up to optimized Triton kernels.
-
AutoMegaKernel
AutoMegaKernel PublicAn agent harness that compiles a model into one provably-correct, self-retargeting CUDA megakernel and self-tunes it past cuBLAS at batch-1 LLM decode, paper: https://arxiv.org/abs/2606.09682
-
qwen3.5-triton
qwen3.5-triton PublicPure Triton kernels for Qwen3.5-27B inference on NVIDIA B200
Repositories
-
- autoevolve Public
Agent-native evolutionary optimization. Say the goal in english, evolve code toward a measured target with a population of coding agents.
- inference-cost-truth Public
What LLM inference actually costs. 874 verified price rows across 24 providers, 378 GPU rental rates, 395 cited throughput datapoints, and a self-hosting break-even model. Every number carries a source URL, a retrieval date, and passed a mechanical evidence gate. Dated immutable snapshots. CC BY 4.0.
- local-kimi Public
Optimized local serving engine for Kimi-Linear-48B: INT4 quantizer, fused decode kernels for a measured 3.18x, and an OpenAI-compatible server. Ships with k3, a bridge that detects the client per request so Claude Code, Codex, Cline, Aider and opencode all work unchanged.
- runinfra-cli Public
- inkling-turbo Public
Faster attention kernels for serving TML's Inkling model on vLLM. 2.7x over the shipping path on H100, and the only implementation that runs on A100.
- bonsai-turbo Public
Single-launch batch-1 decode engine for PrismML Bonsai 27B (ternary and 1-bit) on NVIDIA GPUs. 1.76x the vendor llama.cpp fork on H100, same outputs.