Pinned Loading

1. federated-learning-lab federated-learning-lab Public

   From-scratch federated learning: FedAvg / FedProx / SCAFFOLD, DP-SGD & secure aggregation, plus FedPer / Byzantine-robust / FedAdam / FedLoRA. 33/33 tests, literature-cross-validated, with honest n…

   Python

2. nccl-collectives-bench nccl-collectives-bench Public

   NCCL collective benchmarks on an 8×H100 NVSwitch host — busbw vs link budget, NVLS/Ring/Tree, small-message latency floors (eager vs CUDA Graph vs symmetric memory), and the TP-decode comms ceiling…

   Python

3. nim-agent-blueprint nim-agent-blueprint Public

   Agentic RAG hallucination evaluation on adversarial SQuAD 2.0 (N=200) — nine gate methods compared (self/cross-family/70B judges, PoLL, CoT, MiniCheck, semantic entropy): grounding beats capacity, …

   Python

4. trtllm-triton-serving trtllm-triton-serving Public

   TensorRT-LLM vs vLLM controlled head-to-head on H100 — 12 studies including a knob-by-knob waterfall reproducing NVIDIA's published 27.7k tok/s (100.3%) and attributing the gap to real serving, plu…

   Python

5. blackwell-tensorcore-kernels blackwell-tensorcore-kernels Public

   Hand-written CUDA Tensor Core GEMM kernels on Blackwell (sm_120) and Hopper (sm_90) — raw mma.sync reaching 106% of the cuBLAS-TC kernel on sm_120, CUTLASS 3.x wgmma at 85.5% of nvjet on H100, and …

   Cuda

6. physgate physgate Public

   Validate LLM-generated robot plans in GPU physics simulation — best-of-N plan selection against the highest-quality verifier (Isaac Lab + ROS2 + MCP). Research prototype.

   Python