Search

SOTA low-bit LLM quantization (INT8/FP8/MXFP8/INT4/MXFP4/NVFP4) & sparsity; leading model compression techniques on PyTorch, TensorFlow, and ONNX Runtime

Codec-Aligned Sparsity as a Foundational Principle for Multimodal Intelligence

SparkDiffusion: Accelerating DiT video generation by 265× via joint Sparsity, Distillation, Quantization, etc.