SOTA 3-bit quants (llama/5196)

gguf : fix comparison (ggml/715)

gguf : add input validation, prevent integer overflows (ggml/709)

ggml : add Vulkan backend (llama/2059)

ggml : add unified SYCL backend for Intel GPUs (llama/2690)

ggml : minor type fix (int64_t -> size_t)

Add OpenCL add kernel (llama/5151)

ggml : update softmax n_task calculation (llama/5126)

minor : clean-up some warnings and style (llama/5094)

ggml : parallelize FP32 conversion when using BLAS (llama/5045)

llava : MobileVLM support (llama/4954)

ggml : check ggml_add src1 type (ggml/708)

ggml : add IQ2 to test-backend-ops + refactoring (llama/4990)

imatrix : offload to GPU support (llama/4957)

ggml : importance matrix support for legacy quants (llama/4969)

ggml : introduce GGML_CALL function annotation (llama/4850)

Add ability to use importance matrix for all k-quants (llama/4930)

2-bit quantizations (llama/4897)

ggml: cache sin/cos for RoPE (llama/4908)

gguf : fix potential infinite for-loop (llama/4600)

llama : ggml-backend integration (llama/4766)

Importance Matrix calculation (llama/4861)

ggml : SOTA 2-bit quants (add IQ2_XS) (llama/4856)

ggml : remove ggml_cpy_inplace and ggml_cont_inplace (ggml/693)

Fix execlp call (ggml/689)

SOTA 2-bit quants (llama/4773)

ggml : do not sched_yield when calling BLAS (llama/4761)

ggml : add ggml_cpu_has_avx_vnni() (llama/4589)

ggml : extend ggml_get_rows, ggml_repeat, ggml_concat (ggml/639)

sync : ggml (VMM, sync-ggml-am, dotprod ARM fixes, CUDA fixes) (#1691)

sync : ggml (ggml_scale, ggml_row_size, etc.) (#1677)

sync : ggml (Metal fixes, new ops, tests) (#1633)

sync : ggml (new ops, new backend, etc) (#1602)

ggml : re-enable blas for src0 != F32 (#1583)

sync : ggml (ggml-alloc + linker + gguf fixes) (#1501)

whisper : add full CUDA and Metal offloading (#1472)

ggml : fix MIN / MAX macro re-definition

sync : ggml (backend v2, k-quants, CUDA opts, Metal opts, etc.) (#1422)

sync : ggml (const correctness)

metal : add F32 support + update bench output

whisper : Metal and ggml-alloc support (#1270)

whisper : fix bench regression + fix performance when using CPU BLAS (#1275)

sync : ggml (HBM + Metal + style) (#1264)

build : do not use _GNU_SOURCE gratuitously (#1129)

ggml : posixify pagesize (#1251)

ggml : sync latest llama.cpp (view_src + alloc improvements) (#1247)

ggml : sync (ggml-alloc, GPU, eps, etc.) (#1220)

ggml : fix compilation errors incurred by -Werror (#1227)

ggml : fix compiling when SSE3 is available but not SSSE3 (#1210)

ggml : detect SSSE3 (#1211)