ggml-org/whisper.cpp

Pull Requests Commits

whisper : bump version to 1.9.4

danbev committed 10 days ago

Verified 7eefef89

ci : add Windows On ARM support to release job (#4048)

danbev committed 10 days ago

Verified 6fb4cd67

whisper : call encoder_begin_callback before language auto-detect (#3936)

javierdejesusda committed 12 days ago

Verified c44b60b8

whisper : re-seed decoder 0 between calls (#4025)

davidcabcab committed 12 days ago

Verified 6d0ed914

ci : update close-issue job to not close issues (#4045)

danbev committed 12 days ago

Verified cec4dbe5

tests : load backends before init when built with GGML_BACKEND_DL (#4031)

apollo-2006 committed 13 days ago

Verified 79f2d921

server : return language in detect response (#4035)

turicas committed 13 days ago

Verified 61e6ccad

ggerganov committed 16 days ago

Verified 52a939a2

ggml : bump version to 0.23.0 (ggml/1618)

ggerganov committed 16 days ago

Verified a937f4e8

metal : add remaining fa-vec tunings for M3 Max (llama/28373)

nikwen committed 16 days ago

Verified 11d4eec8

ggml : replace compile definitions with version.h.in (llama/28364)

danbev committed 16 days ago

Verified 140e57a4

ggml : rename and make private ggml_op_alloc_size_may_expand() (ggml/0)

ggerganov committed 16 days ago

Verified e2389eb9

ggml : don't crash when backend search path can't be read (llama/28271)

angt committed 16 days ago

Verified 1b37bead

ggml : remove GGML_CUDA_PEER_MAX_BATCH_SIZE (llama/28177)

angt committed 16 days ago

Verified f32e6fa0

ggml-cpu(s390x) : fix q5_1 uninitialized v_acc (llama/28332)

taronaeo committed 16 days ago

Verified e1bbe405

sycl: fuse rms_norm+mul+add and add+add residual chains (llama/27610)

newjordan committed 16 days ago

Verified d1e0e649

SYCL: Refactor GGML_SYCL_ENABLE_MKL_FA to global var (llama/26863)

johnkarlhill committed 16 days ago

Verified 36f170e5

opencl: quant lm_head / decode GEMV and medium-batch GEMM optimizations (speculative decoding/MTP) (llama/26477)

wanghqc committed 16 days ago

Verified d784add7

tune MMVQ to MMQ crossover for SM87 (llama/28285)

kbenkhaled committed 16 days ago

Verified 0a4a95c8

metal : add sparse FA (llama/28098)

ggerganov committed 16 days ago

Verified 4dd48dde

metal : fix glu dispatch with ne00 = 1 (llama/28306)

ggerganov committed 16 days ago

Verified d55d345e

CUDA: Allow concurrent streams per split for multi-GPU (llama/28198)

tannerbruhn committed 16 days ago

Verified 25350b57

vulkan: fix FA dequant path engagement (llama/28190)

Nathanw1014 committed 16 days ago

Verified 47d348a2

sycl : enhance the api to support peer-to-peer copy (llama/27550)

arthw committed 16 days ago

Verified f24a3860

sycl: reduce redundant work in Q4_K multi-column MMVQ (llama/27062)

Eurekatic committed 16 days ago

Verified a704770e

finetune: fix no KV cache (llama/27199)

ngxson committed 16 days ago

Verified e5605697

ggml-hexagon: add F16 support for unary ops (llama/28228)

cqderek committed 16 days ago

Verified 37f0f443

metal : add fa-vec tunings for M3 (llama/28236)

init-22 committed 16 days ago

Verified 1bdda1e3

metal : fix memory query under low-memory conditions (llama/27701)

madsmtm committed 16 days ago

Verified 3a1c7d6b

ggml-cuda : remove unused vars (llama/28235)

angt committed 16 days ago

Verified 4d343d7c

Older