Popular repositories Loading

1. vllm-gb10-gemma4 vllm-gb10-gemma4 Public

   Complete vLLM + Gemma 4 for NVIDIA DGX Spark GB10 — one command install with benchmarks

   Shell 1 1

2. manthanquant manthanquant Public

   3-bit Lloyd-Max KV Cache Compression for LLM Inference on NVIDIA DGX Spark GB10 — 5.12x compression, 0.983 cosine similarity, pure numpy on ARM unified memory

   Python 1

3. vllm-gb10 vllm-gb10 Public

   Custom native vLLM for NVIDIA DGX Spark GB10 (ARM aarch64, Blackwell sm_121)

   Shell 1

4. vllm-gemma4-patch vllm-gemma4-patch Public

   Gemma 4 support patch for vLLM 0.18.x — backports PR #38826

   Shell 1

5. manthanquant-x86 manthanquant-x86 Public

   TurboQuant KV Cache Compression for vLLM on x86 GPUs — 5.12x compression, 0.983 cosine similarity (BiltIQ AI)

   Python