b5299

llama : support tie embedding for chatglm models (#13328)

b5298

CUDA: mix virt/real CUDA archs for GGML_NATIVE=OFF (#13135)

b5297

clip : refactor graph builder (#13321)

* mtmd : refactor graph builder

* fix qwen2vl

* clean up siglip cgraph

* pixtral migrated

* move minicpmv to a dedicated build function

* move max_feature_layer to build_llava

* use build_attn for minicpm resampler

* fix windows build

* add comment for batch_size

* also support tinygemma3 test model

* qwen2vl does not use RMS norm

* fix qwen2vl norm (2)

b5296

sampling : make top_n_sigma no-op at <=0 or a single candidate (#13345)

b5295

sampling : don't consider -infinity values in top_n_sigma (#13344)

b5293

SYCL: Disable reorder optimize by default and stop setting tensor ext…

b5292

llama : fix build_ffn without gate (#13336)

* llama : fix build_ffn without gate

* fix build on windows

* Revert "fix build on windows"

This reverts commit fc420d3c7eef3481d3d2f313fef2757cb33a7c56.

b5289

CUDA: fix --split-mode row for MMQ (#13323)

b5287

CUDA: fix logic for clearing padding with -ngl 0 (#13320)

b5286

sampling : Integrate Top-nσ into main sampling chain (and add it to t…