DynaTE: Accelerating Diffusion LLMs via Dynamic Token Execution
Diffusion-based LLMs (dLLMs) have recently emerged as a promising alternative to autoregressive (AR) LLMs by enabling bidirectional parallel refinement, alleviating the sequential decoding bottleneck of AR generation.
Key points
- However, their parallel iterative refinement mismatches AR accelerators optimized for sequential decoding and their discrete token generation differs from DiT accelerators designed for continuous denoising.
- Recent dLLM accelerators have explored workload-specific optimizations to reduce vocabulary processing overhead and redundant computation across denoising iterations.
- This paper presents DynaTE, a hardware--software co-design architecture that dynamically adapts accelerator execution to evolving token states during dLLM decoding.
- Second, DynaTE exploits dynamic token dependencies through FLDD to refine a small number of locally dependent tokens within the current iteration, reducing the overall number of denoising iterations, while a Merge--Split--Merge dataflow hides the resulting serial overhead.
Sources (1)
- [1]DynaTE: Accelerating Diffusion LLMs via Dynamic Token ExecutionarXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 8, 05:49 AM
Diffusion-based LLMs (dLLMs) have recently emerged as a promising alternative to autoregressive (AR) LLMs by enabling bidirectional parallel refinement, alleviating the sequential decoding bottleneck of AR generation.
However, their parallel iterative refinement mismatches AR accelerators optimized for sequential decoding and their discrete token generation differs from DiT accelerators designed for continuous denoising.
Extractive summary: sentences quoted from the sources.