@pythongiant(github) is an AI/ML researcher and engineer, currently a Research Intern at Microsoft Research India/ I build open-source tools and models around LLMs, transformers, and inference systems.
KVBoost is a drop-in replacement for AutoModelForCausalLM. Same API surface (KVBoost.from_pretrained(...), engine.generate(...)), but with cross-request KV reuse, FlashAttention-2, AWQ layer streaming, and speculative decoding bolted on.
Thanks! this is a weekend project that i am working on in the side just to learn more about ml engineering and custom cuda kernels. didnt think much about the website
my initial choice was to use Rust for this actually (Probably should've too :P) but i went with python for an initial mvp/skeleton for a future rewrite
KVBoost is a chunk-level KV cache reuse library for HuggingFace models (pip install kvboost). It supports two recompute strategies (selective boundary and CacheBlend), int8/int4 KV quantization for 2–4x RAM reduction, disk-backed cold storage, and 11 architectures including Llama, Qwen, Gemma, Mistral, and Phi. On Qwen2.5-3B we measured 47.9x TTFT speedup on an 8-turn conversation, 21x on code context reuse, 100–743x faster than MLX, and 3–41x faster than vLLM-MLX — including interior chunk reuse where vLLM gets zero hits. Outputs are token-for-token identical to baseline under greedy decoding. Works best on 3B+ models with 500+ token shared context. GitHub: https://github.com/pythongiant/KVBoost
just published a beginner-friendly, hands-on guide to GPU programming with CUDA aimed at folks who are comfortable with basic programming but new to parallel computing. Instead of throwing API references and jargon at you, this guide takes you through the why and how of CUDA step by step. It starts from first principles and builds up to real, runnable code. The goal is to make CUDA intuitive. you’ll walk away understanding how your GPU executes work and how to write code that actually runs fast.