Spaces:

Duplicated from natasa365/whisper.cpp

Xenobd
/

whisper.cpp

Running

App Files Files Community

14.2 MB

100 contributors

History: 911 commits

JohannesGaessler's picture

JohannesGaessler

CUDA: faster q8_0 -> f16 dequantization (llama/4895)

0a1a178 unverified almost 2 years ago

.devops
docker : fix the publishing of the CUDA Docker image (#1704) about 2 years ago
.github
ci : build with CLBlast + ggml-opencl use GGML_API (#1576) about 2 years ago
bindings
go : add SetInitialPrompt method to bindings (#1753) almost 2 years ago
cmake
cmake : update to 3.19 (#351) almost 3 years ago
coreml
coreml : fix ANE optimized encoder (#1716) about 2 years ago
examples
talk-llama : add optional CLI arg to set the bot name (#1764) almost 2 years ago
extra
sync : ggml almost 2 years ago
grammars
whisper : add grammar-based sampling (#1229) about 2 years ago
models
models : make all scripts to be POSIX Compliant (#1725) almost 2 years ago
openvino
whisper : replace `tensor->n_dims` with `ggml_n_dims(tensor)` (#1694) about 2 years ago
samples
Create README.md about 3 years ago
spm-headers
swift : remove local ggml.h reference almost 2 years ago
tests
whisper : make large version explicit + fix data size units (#1493) about 2 years ago
.gitattributes

804 Bytes

Initial release over 3 years ago
.gitignore

803 Bytes

server : add a REST Whisper server example with OAI-like API (#1380) about 2 years ago
.gitmodules

96 Bytes

cmake : add submodule whisper.spm about 3 years ago
CMakeLists.txt

19 kB

release : v1.5.4 almost 2 years ago
LICENSE

1.07 kB

license : update year (#739) over 2 years ago
Makefile

14.7 kB

sync : ggml (VMM, sync-ggml-am, dotprod ARM fixes, CUDA fixes) (#1691) about 2 years ago
Package.swift

1.78 kB

swift : track ggml release branch almost 2 years ago
README.md

37 kB

release : v1.5.4 almost 2 years ago
ggml-alloc.c

29.4 kB

llama : ggml-backend integration (llama/4766) almost 2 years ago
ggml-alloc.h

4.06 kB

llama : ggml-backend integration (llama/4766) almost 2 years ago
ggml-backend-impl.h

5.03 kB

llama : ggml-backend integration (llama/4766) almost 2 years ago
ggml-backend.c

60.7 kB

backend_sched : fix assignments almost 2 years ago
ggml-backend.h

10.4 kB

llama : ggml-backend integration (llama/4766) almost 2 years ago
ggml-cuda.cu

423 kB

CUDA: faster q8_0 -> f16 dequantization (llama/4895) almost 2 years ago
ggml-cuda.h

1.94 kB

llama : ggml-backend integration (llama/4766) almost 2 years ago
ggml-impl.h

7.57 kB

llama : ggml-backend integration (llama/4766) almost 2 years ago
ggml-metal.h

4.33 kB

ggml : add error handling to graph_compute (#1714) about 2 years ago
ggml-metal.m

152 kB

llama : ggml-backend integration (llama/4766) almost 2 years ago
ggml-metal.metal

214 kB

ggml : SOTA 2-bit quants (add IQ2_XS) (llama/4856) almost 2 years ago
ggml-opencl.cpp

82 kB

llama : ggml-backend integration (llama/4766) almost 2 years ago
ggml-opencl.h

1.27 kB

llama : ggml-backend integration (llama/4766) almost 2 years ago
ggml-quants.c

312 kB

ggml : fix 32-bit ARM compat for IQ2_XS (#1758) almost 2 years ago
ggml-quants.h

11.6 kB

ggml : SOTA 2-bit quants (add IQ2_XS) (llama/4856) almost 2 years ago
ggml.c

657 kB

llama : ggml-backend integration (llama/4766) almost 2 years ago
ggml.h

83.3 kB

llama : ggml-backend integration (llama/4766) almost 2 years ago
whisper.cpp

233 kB

whisper : load the model into multiple buffers of max size 1GB (#1763) almost 2 years ago
whisper.h

30.2 kB

whisper : remove trailing whitespaces about 2 years ago