Commit History

Автор SHA1 Съобщение Дата
  matt23654 909072abcf cuda : fix UMA detection on discrete GPUs. (#17537) преди 2 месеца
  Alberto Cabrera Pérez cd8370b408 ggml-cpu: aarm64: q4_K repack gemm and gemv implementations (dotprod only) (#17494) преди 2 месеца
  Eric Curtin d21a76ac38 devops: Add build-essential to Ubuntu 26.04 image (#17531) преди 2 месеца
  Aleksei Nikiforov 4fcd87cf7c gguf-py : skip endian-conversion of MXFP4 data (#17523) преди 2 месеца
  Acly b78db3bd50 vulkan : move contiguous checks to device_supports_op (#17490) преди 2 месеца
  Jeff Bolz 142df17c9c vulkan: use a fixed 1KB buffer for the add_rms_fusion opt (#17514) преди 2 месеца
  Xuan-Son Nguyen e509411cf1 server: enable jinja by default, update docs (#17524) преди 2 месеца
  lhez 7cba58bbea opencl: add sqr, sqrt, mean and ssm_conv (#17476) преди 2 месеца
  Alberto Cabrera Pérez 5449367b21 Fix chunks being too small with small matrix sizes (#17526) преди 2 месеца
  Han Qingzhe 1d594c295c clip: (minicpmv) fix resampler kq_scale (#17516) преди 2 месеца
  Jeff Bolz eec1e33a9e vulkan: allow graph_optimize for prompt processing workloads (#17475) преди 2 месеца
  Jeff Bolz 879d673759 vulkan: Implement top-k (#17418) преди 2 месеца
  xctan 6ab4e50d9c ggml-cpu : add RISC-V Zvfh impl for ggml_vec_mad_f16 (#17448) преди 2 месеца
  Adrien Gallouët 2336cc4784 cmake : use EXCLUDE_FROM_ALL to avoid patch-boringssl.cmake (#17520) преди 2 месеца
  Adrien Gallouët e6923caaec ggml : fix ARM feature verification (#17519) преди 2 месеца
  Jiacheng (Jason) Chen 3e18dba9fd HIP: Patch failed testcase in WMMA-MMQ kernels for RDNA 4 (#17502) преди 2 месеца
  hipudding eeb5605de2 CANN: Add MROPE and IMROPE support (#17401) преди 2 месеца
  o7si f3a848a3b1 chore: upgrade cpp-httplib from v0.27.0 to v0.28.0 (#17513) преди 2 месеца
  Jeff Bolz b3b03a7baf vulkan: Implement GGML_OP_CUMSUM (#17479) преди 2 месеца
  Georgi Gerganov 583cb83416 ggml : add ggml_top_k (#17365) преди 2 месеца
  Aleksei Nikiforov 05872ac885 convert : fix big-endian conversion (#17431) преди 2 месеца
  Diego Devesa 55ab25caf5 codeowners : remove slaren (#17492) преди 2 месеца
  TianHao324 064c90d843 CANN: supports out_prod operator for F32 and F16 (#17406) преди 2 месеца
  Pascal b1846f1c8e webui: add rehype plugin to restore HTML in Markdown table cells (#17477) преди 2 месеца
  Jeff Bolz d414db02d3 vulkan: Use fewer rows for scalar FA when HS is not a multiple of 16 (#17455) преди 2 месеца
  Aaron Teo 877566d512 llama: introduce support for model-embedded sampling parameters (#17120) преди 2 месеца
  Jeff Bolz 3d07caa99b vulkan: more FA details in vk_perf_logger (#17443) преди 2 месеца
  Daniel Bevenius 134e6940ca llama : skip output reordering for single token batches (#17466) преди 2 месеца
  Jiacheng (Jason) Chen 0543f928a3 HIP: WMMA-MMQ kernels for RDNA 4 (#17156) преди 2 месеца
  Sigbjørn Skjæret b61de2b2df convert : allow quantizing lora again (#17453) преди 2 месеца