b4061

metal : reorder write loop in mul mat kernel + style (#10231)

* metal : reorder write loop

* metal : int -> short, style

ggml-ci

b4060

metal : fix build and some more comments (#10229)

b4059

metal : fix F32 accumulation in FA vec kernel (#10232)

b4058

llama : fix Qwen model type strings

b4057

metal : hide debug messages from normal log

b4056

ggml: fix zero division in ‘dne’ calculation in CUDA COUNT_EQUAL oper…

b4055

ggml : optimize llamafile cpu matrix multiplication for ppc64le (#10156)

This change upstreams llamafile's cpu matrix
multiplication kernels for ppc64le using MMA
builtins for FP32 datatype.

This change results in a consistent 90%
improvement in input processing time, and 20%
to 80% improvement in output processing time,
across various batch sizes.

The patch is tested with Meta-Lllama-3-8B,
Mistral-7B, Llama-2-7B-chat-hf models on a
IBM POWER10 machine.

Signed-off-by: Amrita H S <amritahs@linux.vnet.ibm.com>

b4053

metal : opt-in compile flag for BF16 (#10218)

* metal : opt-in compile flag for BF16

ggml-ci

* ci : use BF16

ggml-ci

* swift : switch back to v12

* metal : has_float -> use_float

ggml-ci

* metal : fix BF16 check in MSL

ggml-ci

b4052

metal : improve clarity (minor) (#10171)

b4050

swift : exclude ggml-metal-embed.metal (#10211)

* llama.swift : exclude ggml-metal-embed.metal

* swift : exclude build/