All tags

Topic: "hybrid-architecture"

    MiniMax 2.7: GLM-5 at 1/3 cost SOTA Open Model
    NVIDIA Nemotron 3: hybrid Mamba-Transformer completely open source models from 30B to 500B
    Qwen3-Next-80B-A3B-Base: Towards Ultimate Training & Inference Efficiency
    Mixtral 8x22B Instruct sparks efficiency memes