ggml: Add initial WebGPU backend (llama/14521)

ggml : initial zDNN backend (llama/14975)

ggml-quants : fix make_qp_quants NANs and IQ1 assertion errors (llama/15379)

vulkan: disable spirv-opt for bfloat16 shaders (llama/15352)

vulkan: Use larger workgroups for mul_mat_vec when M is small (llama/15355)

vulkan: support sqrt (llama/15370)

vulkan: Optimize argsort (llama/15354)

vulkan: fuse adds (llama/15252)

vulkan: Support mul_mat_id with f32 accumulators (llama/15337)

vulkan: Add missing bounds checking to scalar/coopmat1 mul_mat_id (llama/15334)

OpenCL: add initial FA support (llama/14987)

opencl: add initial mxfp4 support via mv (llama/15270)

vulkan : fix out-of-bounds access in argmax kernel (llama/15342)

vulkan : fix compile warnings on macos (llama/15340)

ggml: initial IBM zDNN backend (llama/14975)

CUDA: fix negative KV_max values in FA (llama/15321)

HIP: Cleanup hipification header (llama/15285)

vulkan: perf_logger improvements (llama/15246)

ggml: fix ggml_conv_1d_dw bug (ggml/1323)

cuda : fix GGML_CUDA_GRAPHS=OFF (llama/15300)

finetune: SGD optimizer, more CLI args (llama/13873)

HIP: bump requirement to rocm 6.1 (llama/15296)

ggml : update `ggml_rope_multi` (llama/12665)

ggml : repack block_iq4_nlx8 (llama/14904)

CUDA: Optimize `reduce_rows_f32` kernel, leading up to 25x perf improvement on kernel-level and 10% perf increase for Gemma3n (llama/15132)

ggml-rpc: chunk send()/recv() to avoid EINVAL for very large tensors over RPC (macOS & others) (llama/15188)

HIP: disable sync warp shuffel operators from clr amd_warp_sync_functions.h (llama/15273)

sycl: Fix and disable more configurations of mul_mat (llama/15151)

opencl: allow mixed f16/f32 `add` (llama/15140)

CUDA cmake: add `-lineinfo` for easier debug (llama/15260)

CANN: GGML_OP_CPY optimization (llama/15070)

musa: fix failures in test-backend-ops for mul_mat_id op (llama/15236)

CANN: Add broadcast for softmax and FA (llama/15208)

kleidiai: fix unsigned overflow bug (llama/15150)

cuda: refactored ssm_scan and use CUB (llama/13291)

CUDA: add attention sinks for tile and wmma (llama/15178)

gguf-py : add Numpy MXFP4 de/quantization support (llama/15111)

ggml : fix field name when new ggml_backend (llama/14944)

CUDA: attention sinks for mma FlashAttention (llama/15157)

opencl: support sink in `soft_max` (attn sinks) (llama/15152)

vulkan: support fattn sinks (llama/15126)

vulkan: Add env var to disable host visible vidmem (llama/15109)

HIP: add cmake option to enable compiler output of kernel resource usage metrics (llama/15103)

CUDA: GEMM for FP32/FP16/BF16 and ne11 <= 16 (llama/15131)

fix profiling crash (llama/15072)

opencl: add `swiglu_oai` and `add_id` (llama/15121)

ggml : fix fallback to CPU for ununsupported ops (llama/15118)

CANN: add support for ACL Graph (llama/15065)

llama : add gpt-oss (llama/15091)

sycl: fix mul_mat selection (llama/15092)