Pinned Loading

1. llm-serving-engine llm-serving-engine Public

   PyTorch LLM serving engine with paged KV caching, continuous batching, Triton kernels, and retained T4 benchmark evidence.

   Python

2. Distributed-API-Gateway Distributed-API-Gateway Public

   Go API gateway with Redis-backed rate limits, caching, circuit breakers, and reproducible k6 workloads.

   Go

3. streaming-lakehouse-pipeline streaming-lakehouse-pipeline Public

   Postgres CDC lakehouse with Kafka, Spark, and Iceberg; replay-safe merges, backfill drills, and local benchmarks.

   Python

4. threadwave threadwave Public

   Go collaboration engine built on ygo; incremental-sync and encoding optimizations, Yjs compatibility, and reproducible microbenchmarks.

   Go

5. benchseed benchseed Public

   Rust TPC-H/TPC-DS data generation built on tpchgen-rs, with safer CSV delimiters and focused allocation and coordination improvements.

   Rust

6. sdxl-skin-tone-steering sdxl-skin-tone-steering Public

   Controlled SDXL image-colour steering study with frozen protocols, retained evidence, and independent-seed replication.

   Python