ggerganov/whisper.cpp

Pull Requests Commits

ggml : fixes after sync (ggml/983)

slaren committed 1 year ago

40e52a76

ggml-backend : add device and backend reg interfaces (llama/9707)

slaren committed 1 year ago

cf977670

Fixed dequant precision issues in Q4_1 and Q5_1 (llama/9711)

OuadiElfarouki committed 1 year ago

df2c364d

ggml-backend : add device and backend reg interfaces (llama/9707)

slaren committed 1 year ago

1acfadb7

Initial cmake support of SYCL for AMD GPUs (llama/9658)

Alberto Cabrera Pérez committed 1 year ago

ea642144

vulkan : do not use tensor->extra (llama/9407)

rgerganov committed 1 year ago

282a8654

ggml/ex: calculate accuracy in graph, adapt MNIST (ggml/980)

JohannesGaessler committed 1 year ago

936cf3be

ggml: refactor cross entropy loss CPU impl. (ggml/976)

JohannesGaessler committed 1 year ago

bc92c2f8

scripts : sync ggml-backend.cpp

ggerganov committed 1 year ago

f7d55e06

whisper : fix excessive memory usage (#2443)

ggerganov committed 1 year ago

Verified f62a546e

examples : update dr_wav.h to newer version (#2449)

RahulVadhyar committed 1 year ago

Verified 2944cb72

talk-llama : sync llama.cpp

ggerganov committed 1 year ago

ccc25472

metal : reduce command encoding overhead (llama/9698)

ggerganov committed 1 year ago

162a4554

ggerganov committed 1 year ago

ff2cb081

test: fix OPT_STEP_ADAMW for test-backend-ops (ggml/974)

JohannesGaessler committed 1 year ago

5e9d6baa

vulkan : mul_mat: fix UB with small warps (ggml/952)

smeso committed 1 year ago

845f8d66

ggml : fix ggml_cast (ggml/973)

iboB committed 1 year ago

31fdf05f

ggml: fix gradient allocation logic (ggml/966)

JohannesGaessler committed 1 year ago

0ac6666c

ggml : define missing HWCAP flags (llama/9684)

ggerganov committed 1 year ago

6c91da80

ggml : add run-time detection of neon, i8mm and sve (llama/9331)

eddnjjn committed 1 year ago

c245168b

Enable use to the rebar feature to upload buffers to the device. (llama/9251)

mtavenrath committed 1 year ago

280fee8f

mtgpu: enable VMM (llama/9597)

yeahdongcn committed 1 year ago

78b4c1c2

ggml : remove assert for AArch64 GEMV and GEMM Q4 kernels (llama/9217)

chaxu01 committed 1 year ago

1edea2eb

cann: fix crash when llama-bench is running on multiple cann devices (llama/9627)

bachelor-dou committed 1 year ago

96808786

CUDA: remove bad assert (ggml/972)

JohannesGaessler committed 1 year ago

bb57ecb8

vulkan : multithread pipeline creation (ggml/963)

jeffbolznv committed 1 year ago

abdb73c7

vulkan : fix build for GGML_VULKAN_RUN_TESTS, add TFLOPS to log (ggml/961)

jeffbolznv committed 1 year ago

391e548a

vulkan : argsort barriers must be under uniform control flow (ggml/951)

smeso committed 1 year ago

2a29afd4

ggml : fix GGML_MAX_N_THREADS + improve formatting (ggml/969)

ggerganov committed 1 year ago

5963004f

server : ffmpeg overwrite leftover temp file (#2431)

gilbertgong committed 1 year ago

Verified ede1718f

Newer Older