Jeff Bolz ea3ebd8b0d vulkan: Fuse mul_mat_id+add_id+mul and mul_mat+add+add. (llama/17287) ...

These both show up in gpt-oss. Also, cleanup the mul_mat_vec fusion code a bit.

2025-11-17 21:05:46 +02:00

..

ggml-blas

…

ggml-cann

CANN: Add cross_entropy_loss op support (llama/16886)

2025-11-17 21:05:46 +02:00

ggml-cpu

ggml-cpu: handle 3d tensors in repack mat_mul (llama/17241)

2025-11-17 21:05:46 +02:00

ggml-cuda

ggml : add ops SOFTPLUS, EXPM1, TRI, SOLVE_TRI, CUMSUM (llama/17063)

2025-11-17 21:05:46 +02:00

ggml-hexagon

hexagon: various Op fixes (llama/17135)

2025-11-17 21:05:46 +02:00

ggml-hip

HIP: fix AMDGPU_TARGETS, update documentation (llama/16803)

2025-11-09 23:38:03 +02:00

ggml-metal

metal : support argsort for ne00 > 1024 (llama/17247)

2025-11-17 21:05:46 +02:00

ggml-musa

…

ggml-opencl

opencl: add fastdiv and use it in set_rows, ported from cuda (llama/17090)

2025-11-17 21:05:46 +02:00

ggml-rpc

rpc : report actual free memory (llama/16616)

2025-10-22 12:58:11 +03:00

ggml-sycl

fix ci crash about SSM_CONV (llama/17169)

2025-11-17 21:05:46 +02:00

ggml-vulkan

vulkan: Fuse mul_mat_id+add_id+mul and mul_mat+add+add. (llama/17287)

2025-11-17 21:05:46 +02:00

ggml-webgpu

ggml webgpu: faster matrix multiplication/matrix-vector multiplication (llama/17031)

2025-11-09 23:38:03 +02:00

ggml-zdnn

…

CMakeLists.txt

cmake : add version to all shared object files (llama/17091)

2025-11-17 21:05:46 +02:00

ggml-alloc.c

ggml-alloc : make gallocr prefer chunks that allow memory reuse (llama/16788)

2025-11-09 23:38:03 +02:00

ggml-backend-impl.h

…

ggml-backend-reg.cpp

Add experimental ggml-hexagon backend for the Hexagon NPU (llama/16547)

2025-11-09 23:38:03 +02:00

ggml-backend.cpp

sched : fix reserve ignoring user tensor assignments (llama/17232)

2025-11-17 21:05:46 +02:00

ggml-common.h

…

ggml-impl.h

ggml : add ops SOFTPLUS, EXPM1, TRI, SOLVE_TRI, CUMSUM (llama/17063)

2025-11-17 21:05:46 +02:00

ggml-opt.cpp

…

ggml-quants.c

…

ggml-quants.h

…

ggml-threading.cpp

…

ggml-threading.h

…

ggml.c

ggml : add ops SOFTPLUS, EXPM1, TRI, SOLVE_TRI, CUMSUM (llama/17063)

2025-11-17 21:05:46 +02:00

ggml.cpp

…

gguf.cpp

…