server : add request path option(#1741)

main : add cli option to disable system prints (#1740)

server : fix server temperature + add temperature_inc (#1729)

talk-llama : sync latest llama.cpp

release : v1.5.4

fix : cuda order of synchronization when setting a buffer (ggml/679)

metal : switch back to default.metallib (ggml/681)

ggml : fix q2_k bpw in comments (ggml/680)

coreml : fix ANE optimized encoder (#1716)

whisper.swiftui : add .gitignore

whispser : reset the "batched" timings (#1721)

release : v1.5.3

swift : update Package.swift to use ggml as package dependency (#1701)

ggml : add error handling to graph_compute (#1714)

cuda : simplify expression

cuda : mark I16 and I32 ops as unsupported

metal : add kernel_get_rows_i32

metal : optimize ggml_mul_mat_id (faster Mixtral PP) (llama/4725)

metal : enable shader debugging (cmake option) (llama/4705)

ggml : add ggml_vdotq_s32 alias (llama/4715)

CUDA: fixed tensor cores not being used on RDNA3 (llama/4697)

ggml : add ggml_cpu_has_avx_vnni() (llama/4589)

CUDA: fix tensor core logic for Pascal and HIP (llama/4682)

cuda: fix vmm oom issue on NVIDIA AGX Orin (llama/4687)

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

scripts : fix sync order + metal sed

examples : fix WASM Stack Overflow (#1713)

docker : fix the publishing of the CUDA Docker image (#1704)

scripts : do not sync commits from this repo

ci : build with CLBlast + ggml-opencl use GGML_API (#1576)

whisper : replace `tensor->n_dims` with `ggml_n_dims(tensor)` (#1694)

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

download : fix large q5 model name (#1695)

whisper : Replace WHISPER_PRINT_DEBUG with WHISPER_LOG_DEBUG (#1681)

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

docker : Dockerize whisper.cpp (#1674)

CI : Add coverage for talk-llama when WHISPER_CUBLAS=1 (#1672)

examples : Revert CMakeLists.txt for talk-llama (#1669)

cmake : set default CUDA architectures (#1667)

bench.py : add different large models (#1655)

wchess : update README.md

release : v1.5.2

wchess : update readme

wchess : whisper assisted chess (#1595)

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

cmake : target windows 8 or above for prefetchVirtualMemory in llama-talk (#1617)

cmake : Fix bug in httplib.h for mingw (#1615)

metal : fix `ggml_metal_log` vargs (#1606)

whisper.objc : disable timestamps for real-time transcription

whisper : more debug messages + fix fallback logic