sync : ggml#2737

Merged

ggerganov merged 23 commits intomasterfrom

sync-ggml-25-01-14

Jan 14, 2025

Commits on Jan 14, 2025

ggml : allow loading backend with env variable (ggml/1059)

rgerganov
authored and
ggerganov
committed
fix: Vulkan shader gen binary path (llama/11037)

giladgd
authored and
ggerganov
committed
Support for models with non-512-aligned tensors over RPC. (llama/11047)

authored and
ggerganov
committed
Vulkan: Add device-specific blacklist for coopmat for the AMD proprietary driver (llama/11074)

0cc4m
authored and
ggerganov
committed
CUDA: add BF16 support (llama/11093)

JohannesGaessler
authored and
ggerganov
committed
SYCL: Use get_multi_ptr instead of deprecated get_pointer in wkv6 (llama/11087)

qnixsynapse
authored and
ggerganov
committed
rpc : code cleanup (llama/11107)

rgerganov
authored and
ggerganov
committed
ggml-backend : only offload from host buffers (llama/11120)

slaren
authored and
ggerganov
committed
ggml-backend : only offload from host buffers (fix) (llama/11124)

slaren
authored and
ggerganov
committed
GGUF: C++ refactor, backend support, misc fixes (llama/11030)

JohannesGaessler
authored and
ggerganov
committed
fix: Vulkan shader gen binary path when Cross-compiling (llama/11096)

ag2s20150909
authored and
ggerganov
committed
Disable GL_KHR_cooperative_matrix Vulkan extension if not available. (llama/11117)

mbaudier
authored and
ggerganov
committed
llamafile : ppc64le MMA INT8 implementation (llama/10912)

amritahs-ibm
authored and
ggerganov
committed
fix: add missing msg in static_assert (llama/11143)

hydai
authored and
ggerganov
committed
SYCL: Refactor ggml_sycl_compute_forward (llama/11121)

qnixsynapse
authored and
ggerganov
committed
llama: add support for QRWKV6 model architecture (llama/11001)

committed
Vulkan: Fix float16 use on devices without float16 support + fix subgroup_size_control validation error (llama/11161)

0cc4m
authored and
ggerganov
committed
ggml : do not define GGML_USE_CUDA when building with GGML_BACKEND_DL (llama/11211)

rgerganov
authored and
ggerganov
committed
cuda : CUDA Graph Compute Function Refactor (precursor for performance improvements) (llama/11042)

authored and
ggerganov
committed
ggml : add opencl backend (skip) (llama/10693)

authored and
ggerganov
committed
GGUF: C++ refactor, backend support, misc fixes (skip) (llama/11030)

JohannesGaessler
authored and
ggerganov
committed
sync : ggml
ggerganov
committed
talk-llama : sync llama.cpp
ggerganov
committed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

sync : ggml#2737

sync : ggml#2737
ggerganov merged 23 commits intomasterfrom
sync-ggml-25-01-14

Commits on Jan 14, 2025

ggml : allow loading backend with env variable (ggml/1059)

fix: Vulkan shader gen binary path (llama/11037)

Support for models with non-512-aligned tensors over RPC. (llama/11047)

Vulkan: Add device-specific blacklist for coopmat for the AMD proprietary driver (llama/11074)

CUDA: add BF16 support (llama/11093)

SYCL: Use get_multi_ptr instead of deprecated get_pointer in wkv6 (llama/11087)

rpc : code cleanup (llama/11107)

ggml-backend : only offload from host buffers (llama/11120)

ggml-backend : only offload from host buffers (fix) (llama/11124)

GGUF: C++ refactor, backend support, misc fixes (llama/11030)

fix: Vulkan shader gen binary path when Cross-compiling (llama/11096)

Disable GL_KHR_cooperative_matrix Vulkan extension if not available. (llama/11117)

llamafile : ppc64le MMA INT8 implementation (llama/10912)

fix: add missing msg in static_assert (llama/11143)

SYCL: Refactor ggml_sycl_compute_forward (llama/11121)

llama: add support for QRWKV6 model architecture (llama/11001)

Vulkan: Fix float16 use on devices without float16 support + fix subgroup_size_control validation error (llama/11161)

ggml : do not define GGML_USE_CUDA when building with GGML_BACKEND_DL (llama/11211)

cuda : CUDA Graph Compute Function Refactor (precursor for performance improvements) (llama/11042)

ggml : add opencl backend (skip) (llama/10693)

GGUF: C++ refactor, backend support, misc fixes (skip) (llama/11030)

sync : ggml

talk-llama : sync llama.cpp

sync : ggml#2737ggerganov merged 23 commits intomasterggml-org/whisper.cpp:masterfrom sync-ggml-25-01-14ggml-org/whisper.cpp:sync-ggml-25-01-14Copy head branch name to clipboard

Commits

Commits on Jan 14, 2025

sync : ggml#2737
ggerganov merged 23 commits intomasterfrom
sync-ggml-25-01-14