AK shares a paper on dMoE model
AK(@_akhaliq)39 字 (约 1 分钟)
85
AK shares a paper on the dMoE model, which uses learnable block experts to implement dLLMs and has better performance and efficiency.
入选理由:dMoE模型使用可学习的块专家来实现dLLMs,具有更好的性能和效率。
FeaturedTweet#dMoE#dLLMs#learnable block experts中文