llama: avoid copying logits during prompt decode in MTP by am17an · Pull Request

<table> <tr><td> <a href="https://www.reddit.com/r/LocalLLaMA/comments/1svdjfa/cuda_reduce_mmq_streamk_overhead_by/"> <img alt="CUDA: reduce MMQ stream-k overhead by

<table> <tr><td> <a href="https://www.reddit.com/r/LocalLLaMA/comments/1teryn8/llama_spec_mtp_support_by_am17an_pull_request/"> <img alt="llama + spec: MTP Support by am17an ·

<table> <tr><td> <a href="https://www.reddit.com/r/LocalLLaMA/comments/1tawn19/feat_add_mimo_v25_vision_by_aessedai_pull_request/"> <img alt="feat: add MiMo v2.5 vision by

<table> <tr><td> <a href="https://www.reddit.com/r/LocalLLaMA/comments/1x6mlby/add_nemotron_super_35_vl_and_nemotron_bolt_74b/"> <img alt="Add Nemotron Super 3.5 VL and Nemotron

<table> <tr><td> <a href="https://www.reddit.com/r/LocalLLaMA/comments/1sn2hjg/ggml_add_graph_reused_by_am17an_pull_request/"> <img alt="ggml: add graph_reused by am17an · Pull

<!-- SC_OFF --><div class="md"><p>PR #22673 (commit 4f13cb7) landed MTP speculative decoding in mainline llama.cpp on May 16. I tested it on two separate rigs.</p> <p>Qwen3.6 27B,

<table> <tr><td> <a href="https://www.reddit.com/r/LocalLLaMA/comments/1uoxdp8/ggmlcpu_use_ue4m3_lut_in_arm_nvfp4_dot_product_by/"> <img alt="ggml-cpu: use UE4M3 LUT in ARM NVFP4

<table> <tr><td> <a href="https://www.reddit.com/r/LocalLLaMA/comments/1syfenw/ggmlcuda_add_flashattn_support_for_dkq320dv256/"> <img alt="ggml-cuda: add flash-attn support for