Tag
Introduces a semi-autoregressive framework that combines latent block diffusion with temporal point processes for generating asynchronous event sequences, reducing error accumulation while enabling variable-length output.
New research on DFlash and Spec V2 speculative decoding methods achieves >4.3X baseline throughput for LLM inference, released as the default speculative decoding engine in SGLang.
DFlash, a block-diffusion drafter with KV injection, is now running at frontier scale, achieving up to 4.3x greater throughput over baseline, integrated with Modal and SGLang for Qwen 397B.
Fast-dDrive is a block-diffusion VLA model for end-to-end autonomous driving that achieves state-of-the-art trajectory accuracy while delivering over 12x throughput speedup over autoregressive baselines, addressing the trade-off between high-fidelity planning and efficient inference for edge deployment.
Z-lab released DFlash, a speculative decoding drafter model for Gemma-4-31B-it that uses lightweight block diffusion to draft multiple tokens in parallel, achieving up to 5.8x speedup over autoregressive baseline.
This article introduces Qwen3.6-27B-DFlash, a specialized drafter model for DFlash, a novel speculative decoding method using block diffusion to accelerate inference speed. It provides installation instructions for vLLM and SGLang to enable parallel drafting with the target Qwen3.6-27B model.
z-lab releases DFlash, a speculative decoding drafter that uses a lightweight block-diffusion model to draft 15–16 tokens in parallel, yielding up to 2.9× speedup for Qwen3.6-35B-A3B inference.