CUDA: fix unused warning in mmq.cu (llama/7442)

CUDA: deduplicate mmq code (llama/7397)

rpc : track allocated buffers (llama/7411)

Update SYCL upscale operation (llama/7321)

ggml-opencl, llama: using reserve() if count already known (llama/7272)

ggml : add loongarch lsx and lasx support (llama/6454)

Add provisions for windows support for BF16 code including CMake provision for enabling AVX512_BF16 (llama/7258)

Vulkan Embedding Fix (llama/7360)

ggml : fix another case of quants nans (llama/7387)

ggml: implement quantized KV cache for FA (llama/7372)

cuda : clear error after buffer allocation failure (llama/7376)

Capture CUDA logging output (llama/7298)

android : use "ci-android" branch for CI (llama/7341)

CUDA: deduplicate FlashAttention code (llama/7352)

cuda : add half2 __shfl_xor() for ROCm 5.5 (llama/7263)

Update and fix Vulkan soft_max and argsort implementations (llama/7237)

ggml : fix quants nans when all the group weights are very close to zero (llama/7313)

CUDA: faster large batch FA without tensor cores (llama/7314)

rpc : set SO_REUSEADDR for the server socket (llama/7320)

ggml-quants, llama : removed excess checks (llama/7274)

ggml : rewrite silu and softmax for cpu (llama/7154)

rpc : add command line arg for specifying backend memory

Add support for properly optimized Windows ARM64 builds with LLVM and MSVC (llama/7191)

ggml : use dynamic thread scheduling for matrix multiplication (llama/6915)

Avoid unnecessarily disabling CUDA graphs (llama/7302)

ggml : tag ggml_tensor::backend as deprecated (llama/7290)

Add missing " (llama/7303)

ggml : add `ggml_upscale_ext` (ggml/814)

scripts : update sync

whisper : use ggml-cuda in mel calc, set appropriate device (#2236)

cuda : fix HIPBLAS build (#2234)

cuda : fix bounds check for src0 rows in MMVQ kernel (#2231)

ci : fix CUDA builds (#2232)

whisper : auto-grow working areas for mel_calc_cuda (#2227)

whisper : free whisper_mel instances (#2220)

whisper : whisper_state/backend fixes (#2217)

whisper : calculate mel spectrogram directly into a ggml_tensor (#2208)

whisper : add CUDA-specific computation mel spectrograms (#2206)

whisper : remove `speed_up` and `phase_vocoder*` functions (#2198)

readme : add conan badge (#2196)

readme : add install instructions for Conan (#2189)

whisper: use global cache for sin/cos vals and Hann window (#2194)

release : v1.6.2

Revert "whisper : remove extra backend instance (huh?)" (#2182)

server : fix typo (#2181)

ruby : update bindings (#2154)

release : v1.6.1

examples : add support for decoding input with ffmpeg (Linux) (#2133)

node : add flash_attn param (#2170)

ci: Update build.yml to suppress warnings about node.js versions (#2166)