readme : update build instructions

ci : disable CUDA and Android builds

ci : disable Obj-C build + fixes

make : shim cmake

talk-llama : sync llama.cpp

sync : ggml

ggml : add predefined list of CPU backend variants to build (llama/10626)

ggml-cpu : fix HWCAP2_I8MM value (llama/10646)

vulkan: Implement "fast divide" (mul+shift) for unary ops like copy (llama/10642)

SYCL : Move to compile time oneMKL interface backend selection for NVIDIA backend (llama/10584)

Avoid using __fp16 on ARM with old nvcc (llama/10616)

vulkan: optimize and reenable split_k (llama/10637)

ggml: add `GGML_SET` Metal kernel + i32 CPU kernel (ggml/1037)

ggml : add `GGML_PAD_REFLECT_1D` operation (ggml/1034)

files : remove make artifacts

common : fix compile warning

ggml : move AMX to the CPU backend (llama/10570)

metal : small-batch mat-mul kernels (llama/10581)

SYCL: Fix and switch to GGML_LOG system instead of fprintf (llama/10579)

ggml-cpu: replace AArch64 NEON assembly with intrinsics in ggml_gemv_q4_0_4x4_q8_0() (llama/10567)

vulkan: Dynamic subgroup size support for Q6_K mat_vec (llama/10536)

ggml : fix I8MM Q4_1 scaling factor conversion (llama/10562)

ggml-cpu: fix typo in gemv/gemm iq4_nl_4_4 (llama/10580)

sycl : offload of get_rows set to 0 (llama/10432)

sycl : Reroute permuted mul_mats through oneMKL (llama/10408)

CANN: RoPE operator optimization (llama/10563)

vulkan: get the first command buffer submitted sooner (llama/10499)

ggml : remove redundant copyright notice + update authors

ggml : fix row condition for i8mm kernels (llama/10561)

cmake : fix ARM feature detection (llama/10543)

ggml-cpu: support IQ4_NL_4_4 by runtime repack (llama/10541)

kompute : improve backend to pass test_backend_ops (llama/10542)

CANN: Fix SOC_TYPE compile bug (llama/10519)

CANN: ROPE operator optimization (llama/10540)

Add some minimal optimizations for CDNA (llama/10498)

metal : fix group_norm support condition (llama/0)

vulkan: define all quant data structures in types.comp (llama/10440)

vulkan: Handle GPUs with less shared memory (llama/10468)

vulkan: further optimize q5_k mul_mat_vec (llama/10479)

vulkan: skip integer div/mod in get_offsets for batch_idx==0 (llama/10506)

vulkan: optimize Q2_K and Q3_K mul_mat_vec (llama/10459)

mtgpu: Add MUSA_DOCKER_ARCH in Dockerfiles && update cmake and make (llama/10516)

vulkan: fix group_norm (llama/10496)

cmake : enable warnings in llama (llama/10474)

ggml-cpu: cmake add arm64 cpu feature check for macos (llama/10487)

CANN: Improve the Inferencing Performance for Ascend NPU Device (llama/10454)

CANN: RoPE and CANCAT operator optimization (llama/10488)

vulkan: Fix a vulkan-shaders-gen arugment parsing error (llama/10484)

metal : enable mat-vec kernels for bs <= 4 (llama/10491)

llama : accept a list of devices to use to offload a model (llama/10497)