Pinned Loading
-
llmterm
llmterm PublicType llm <prompt> or llm! <prompt> in any terminal — runs your existing Claude Code, Codex, or Gemini CLI as an agent. Zsh widget, no PTY hijack, no extra subscription
Go
-
attestrag
attestrag PublicConfidential RAG: documents, index and generation stay inside SEV-SNP / TDX CVMs with NVIDIA CC GPUs; keys are released only after CPU+GPU attestation (Tamarin-proven release protocol, Lean-checked…
Python
-
abyss
abyss PublicRun DeepSeek-V4.1-Flash (763B, native MXFP4) on one two-GPU workstation by tiering weights across VRAM, DRAM and NVMe. 在家用双卡工作站上分层运行 DeepSeek V4.1 Flash(显存、内存、NVMe 三层)
C++
-
build-an-llm-inference-engine
build-an-llm-inference-engine PublicBuild an LLM Inference Engine (From Scratch) · 《从零构建大模型推理引擎》:纯 C + CUDA 实现 Qwen3.8-27B 推理
C 3
If the problem persists, check the GitHub status page or contact support.