Pascal 32560ea597 talker, code-predictor: fused flash attention, fix Vulkan autoregressive decode

2026-05-11 15:29:24 +02:00

examples

example file for voice clone

2026-05-10 17:20:15 +02:00

ggml @ e2196fdc83

ggml: sync from llama.cpp

2026-05-10 20:56:28 +02:00

src

talker, code-predictor: fused flash attention, fix Vulkan autoregressive decode

2026-05-11 15:29:24 +02:00

tests

tests logs

2026-05-11 13:38:21 +02:00

tools

loader, quantize: align GGUF on llama.cpp, conv kernels widened to F16 at load

2026-05-11 12:42:54 +02:00

.clang-format

Initial release

2026-05-10 15:57:15 +02:00

.gitattributes

Initial release

2026-05-10 15:57:15 +02:00

.gitignore

Initial release

2026-05-10 15:57:15 +02:00

.gitmodules

Initial release

2026-05-10 15:57:15 +02:00

buildall.cmd

Initial release

2026-05-10 15:57:15 +02:00

buildall.sh

Initial release

2026-05-10 15:57:15 +02:00

buildcpu.sh

Initial release

2026-05-10 15:57:15 +02:00

buildcuda.cmd

Initial release

2026-05-10 15:57:15 +02:00

buildcuda.sh

Initial release

2026-05-10 15:57:15 +02:00

buildtermux.sh

Initial release

2026-05-10 15:57:15 +02:00

buildvulkan.cmd

Initial release

2026-05-10 15:57:15 +02:00

buildvulkan.sh

Initial release

2026-05-10 15:57:15 +02:00

checkpoints.sh

Initial release

2026-05-10 15:57:15 +02:00

CMakeLists.txt

Initial release

2026-05-10 15:57:15 +02:00

convert.py

convert: drop orphan encoder.sliding_window GGUF key

2026-05-11 05:48:47 +02:00

format.sh

Initial release

2026-05-10 15:57:15 +02:00

LICENSE

Initial release

2026-05-10 15:57:15 +02:00

models.sh

Initial release

2026-05-10 15:57:15 +02:00

quantize.sh

Initial release

2026-05-10 15:57:15 +02:00

update.cmd

Initial release

2026-05-10 15:57:15 +02:00

update.sh

Initial release

2026-05-10 15:57:15 +02:00

S

Description

qwentts.cpp fork: tts-server Qwen3-TTS con containerizzazione Vulkan

MIT

2.9 MiB

Languages

C++ 81.6%

Python 13.1%

C 1.6%

Shell 1.4%

CMake 1.2%

Other 1.1%