b6795

HIP: fix GPU_TARGETS (#16642)

b6791

llama-model: fix insonsistent ctxs <-> bufs order (#16581)

b6782

mtmd : support home-cooked Mistral Small Omni (#14928)

b6775

gguf-py : add support for endian conversion of BF16 data (#16594)

BF16 requires special handling in this script
while it's a 2-bytes data, but view is 1-byte by default.
Switch to correct view before attempting byteswapping.

With this change correctly byteswapping models like
Meta-Llama-3-8B-Instruct-bf16-GGUF
should be possible.

b6773

opencl: add q8_0 mm support (#16469)

* opencl: add mm_q8_0_f32

* opencl: fix data loading for incomplete tile

* opencl: use q8_0 mm for larger matrix

* opencl: add some tests to cover the path

b6772

opencl: fix FA for f32 (#16584)

b6765

metal : avoid using Metal's gpuAddress property (#16576)

* metal : avoid using Metal's gpuAddress property

* metal : fix rope kernels buffer check

b6756

server : dynamic token limit for prompt cache (#16560)

* server : dynamic token limit for prompt cache

* cont : print estimated token limit

b6724

cpu : optimize the ggml NORM operation (#15953)

* ggml-cpu: optimize norm operation to use intrinsics or Accelerate

          rename function

          add endif macro comment

Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>
Co-authored-by: Aaron Teo <taronaeo@gmail.com>

* implement s390x SIMD suggested by @taronaeo

* add TODO comment

* tidy up spaces

---------

Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>
Co-authored-by: Aaron Teo <taronaeo@gmail.com>

b6714

Disable CUDA host buffers on integrated GPUs (#16308)