ggml-org/whisper.cpp

Pull Requests Commits

ggml : disable CUDA graphs for non-llama.cpp projects

ggerganov committed 2 years ago

Verified ceb77363

whisper : disable CUDA mel + fix FFMPEG

ggerganov committed 2 years ago

Verified dc8cc2dd

ggerganov committed 2 years ago

Verified 3efedb95

whisper : reorganize source code + improve CMake (#2256)

ggerganov committed 2 years ago

Verified e30c6799

whisper : optimize fft() function (#2242)

mkycoder committed 2 years ago

Verified bf4cb4ab

talk-llama : sync llama.cpp

ggerganov committed 2 years ago

Verified e293f17d

whisper : use ggml_backend_sched (#2239)

ggerganov committed 2 years ago

5d950c4b

fix : remove extra files

ggerganov committed 2 years ago

820446e2

scripts : sync ggml-blas

ggerganov committed 2 years ago

54d5823e

build : update make / cmake

ggerganov committed 2 years ago

5181494e

ggerganov committed 2 years ago

4a6e6e8b

move BLAS to a separate backend (cont) (llama/6210)

slaren committed 2 years ago

de29b193

Vulkan Shader Refactor, Memory Debugging Option (llama/7947)

0cc4m committed 2 years ago

92297104

scripts : stop sync whisper example from ggml

ggerganov committed 2 years ago

63a767a1

cmake : fix sycl build (#0)

ggerganov committed 2 years ago

30841fa7

ggml : remove OpenCL (#0)

ggerganov committed 2 years ago

3b1ac038

sycl : sync (#0)

ggerganov committed 2 years ago

990de617

cuda : enable CUDA graphs (#0)

ggerganov committed 2 years ago

6975600b

talk-llama : sync llama.cpp

ggerganov committed 2 years ago

061eeb9f

cmake : fix CUDA build (#0)

ggerganov committed 2 years ago

4942b1b4

ggerganov committed 2 years ago

3c7cc5c4

ggml : fix and optimize ppc64le (ggml/849)

penghongbo committed 2 years ago

5cd42ee2

ggml : remove duplicate include of ggml-common.h (ggml/853)

danbev committed 2 years ago

ee718f3d

remove global variables (llama/7710)

airMeng committed 2 years ago

63eac1f6

CUDA: faster q2_K, q3_K MMQ + int8 tensor cores (llama/7921)

JohannesGaessler committed 2 years ago

b17ba281

metal : utilize max shared memory for mul_mat_id (llama/7935)

ggerganov committed 2 years ago

7a489af2

rpc : fix ggml_backend_rpc_supports_buft() (llama/7918)

rgerganov committed 2 years ago

4a4ea13d

move BLAS to a separate backend (llama/6210)

slaren committed 2 years ago

174a461f

CUDA: fix broken oob check for FA vec f32 kernel (llama/7904)

JohannesGaessler committed 2 years ago

d8b7a24b

tests : add non-cont unary tests (llama/7857)

ggerganov committed 2 years ago

acf3832c

Older