GGML update to ec98e2002 (#13451)

* Revert "add support for NVIDIA Nemotron 3 Nano" This reverts commit e7d2ae9d69421012e9a8765c06a3fdf0e45b12f3. * GGML update to 380b4c984 Remove MaskBatchPadding as GGML_KQ_MASK_PAD is no longer present (no padding required) * update to c45f89d55 * ec98e2002 solar pro needed more adjusting - needs verification * review comments

GGML update to ec98e2002 (#13451)
* Revert "add support for NVIDIA Nemotron 3 Nano" This reverts commit e7d2ae9d69421012e9a8765c06a3fdf0e45b12f3. * GGML update to 380b4c984 Remove MaskBatchPadding as GGML_KQ_MASK_PAD is no longer present (no padding required) * update to c45f89d55 * ec98e2002 solar pro needed more adjusting - needs verification * review comments
49a9c9ba · Daniel Hiltgen · GitHub · 1c094038 · 49a9c9ba · 49a9c9ba
Unverified Commit 49a9c9ba authored Dec 17, 2025 by Daniel Hiltgen Committed by GitHub Dec 17, 2025
20 changed files
--- a/llama/patches/0022-ggml-Enable-resetting-backend-devices.patch
+++ b/llama/patches/0022-ggml-Enable-resetting-backend-devices.patch
--- a/llama/patches/0024-GPU-discovery-enhancements.patch
+++ b/llama/patches/0024-GPU-discovery-enhancements.patch
--- a/llama/patches/0028-Add-memory-detection-using-DXGI-PDH.patch
+++ b/llama/patches/0028-Add-memory-detection-using-DXGI-PDH.patch
--- a/llama/patches/0029-ggml-cuda-skip-large-batches.patch
+++ b/llama/patches/0029-ggml-cuda-skip-large-batches.patch
--- a/llama/patches/0030-win-exit-instead-of-abort.patch
+++ b/llama/patches/0030-win-exit-instead-of-abort.patch
--- a/llama/patches/0031-fix-bakllava-regression.patch
+++ b/llama/patches/0031-fix-bakllava-regression.patch
--- a/llama/patches/0032-llama-add-support-for-NVIDIA-Nemotron-Nano-3.patch
+++ b/llama/patches/0032-llama-add-support-for-NVIDIA-Nemotron-Nano-3.patch
--- a/llama/sampling_ext.cpp
+++ b/llama/sampling_ext.cpp
--- a/ml/backend.go
+++ b/ml/backend.go
--- a/ml/backend/ggml/ggml.go
+++ b/ml/backend/ggml/ggml.go
--- a/ml/backend/ggml/ggml/include/ggml-alloc.h
+++ b/ml/backend/ggml/ggml/include/ggml-alloc.h
--- a/ml/backend/ggml/ggml/include/ggml-backend.h
+++ b/ml/backend/ggml/ggml/include/ggml-backend.h
--- a/ml/backend/ggml/ggml/include/ggml-cpu.h
+++ b/ml/backend/ggml/ggml/include/ggml-cpu.h
--- a/ml/backend/ggml/ggml/include/ggml-zendnn.h
+++ b/ml/backend/ggml/ggml/include/ggml-zendnn.h
--- a/ml/backend/ggml/ggml/include/ggml.h
+++ b/ml/backend/ggml/ggml/include/ggml.h
--- a/ml/backend/ggml/ggml/src/ggml-alloc.c
+++ b/ml/backend/ggml/ggml/src/ggml-alloc.c
--- a/ml/backend/ggml/ggml/src/ggml-backend.cpp
+++ b/ml/backend/ggml/ggml/src/ggml-backend.cpp
--- a/ml/backend/ggml/ggml/src/ggml-cpu/arch/arm/repack.cpp
+++ b/ml/backend/ggml/ggml/src/ggml-cpu/arch/arm/repack.cpp
--- a/ml/backend/ggml/ggml/src/ggml-cpu/ggml-cpu.c
+++ b/ml/backend/ggml/ggml/src/ggml-cpu/ggml-cpu.c
--- a/ml/backend/ggml/ggml/src/ggml-cpu/ggml-cpu.cpp
+++ b/ml/backend/ggml/ggml/src/ggml-cpu/ggml-cpu.cpp