Commit History

vulkan : retry allocation with fallback flags (#2451)
9e91cbc
unverified

SRHMorris fdsffdsafds commited on

release : v1.7.0
5abc285
unverified

ggerganov commited on

scripts : bench v3-turbo
22b5adf
unverified

ggerganov commited on

whisper : remove mel leftover constants (396089f)
505ec31
unverified

ggerganov commited on

whisper : zero-out the KV cache upon clear (#2445)
b8af443

ggerganov commited on

objc : fix build
564bce1

ggerganov commited on

metal : zero-init buffer contexts (#0)
d651546

ggerganov commited on

whisper : revert mel-related changes (#0)
7708c26

ggerganov commited on

whisper : adapt to latest ggml (skip) (#0)
ad9dd7b

ggerganov commited on

ggml : fix typo in example usage ggml_gallocr_new (ggml/984)
30a097b

danbev commited on

ggml : fixes after sync (ggml/983)
237c05a

Diego Devesa commited on

ggml-backend : add device and backend reg interfaces (llama/9707)
9d74d85

Diego Devesa commited on

Fixed dequant precision issues in Q4_1 and Q5_1 (llama/9711)
5239c28

Ouadie EL FAROUKI commited on

ggml-backend : add device and backend reg interfaces (llama/9707)
1bdb50a

Diego Devesa JohannesGaessler commited on

Initial cmake support of SYCL for AMD GPUs (llama/9658)
7d7ac98

Alberto Cabrera Pérez commited on

ggml/ex: calculate accuracy in graph, adapt MNIST (ggml/980)
52069b8

JohannesGaessler commited on

ggml: refactor cross entropy loss CPU impl. (ggml/976)
2a0805f

JohannesGaessler commited on

scripts : sync ggml-backend.cpp
26efed1

ggerganov commited on

whisper : fix excessive memory usage (#2443)
afe3785
unverified

ggerganov commited on

examples : update dr_wav.h to newer version (#2449)
d678325
unverified

Rahul Vadhyar commited on

talk-llama : sync llama.cpp
c9ddda2

ggerganov commited on

metal : reduce command encoding overhead (llama/9698)
43d5a06

ggerganov commited on

sync : ggml
c5e24da

ggerganov commited on

test: fix OPT_STEP_ADAMW for test-backend-ops (ggml/974)
76aa810

JohannesGaessler commited on

vulkan : mul_mat: fix UB with small warps (ggml/952)
d1a29c6

smeso commited on

ggml : fix ggml_cast (ggml/973)
c44d575

stanimirovb commited on

ggml: fix gradient allocation logic (ggml/966)
ad3f29d

JohannesGaessler commited on

ggml : define missing HWCAP flags (llama/9684)
1d52105

ggerganov Willy Tarreau commited on

ggml : add run-time detection of neon, i8mm and sve (llama/9331)
12c0e23

Dan Johansson commited on

Enable use to the rebar feature to upload buffers to the device. (llama/9251)
760f8c2

Markus Tavenrath commited on

mtgpu: enable VMM (llama/9597)
e84b4f5

R0CKSTAR commited on

ggml : remove assert for AArch64 GEMV and GEMM Q4 kernels (llama/9217)
50395aa

Charles Xu commited on

cann: fix crash when llama-bench is running on multiple cann devices (llama/9627)
068c697

dou112 commited on

CUDA: remove bad assert (ggml/972)
91954a7

JohannesGaessler commited on

vulkan : multithread pipeline creation (ggml/963)
ba60f98

jeffbolznv commited on

vulkan : fix build for GGML_VULKAN_RUN_TESTS, add TFLOPS to log (ggml/961)
85e2387

jeffbolznv commited on

vulkan : argsort barriers must be under uniform control flow (ggml/951)
b2602d7

smeso commited on

ggml : fix GGML_MAX_N_THREADS + improve formatting (ggml/969)
ad34655

ggerganov commited on

server : ffmpeg overwrite leftover temp file (#2431)
2dafb8e
unverified

dynafire commited on

whisper : add large-v3-turbo (#2440)
f3283ba
unverified

ggerganov commited on

tests : remove test-backend-ops (#2434)
050ba38
unverified

ggerganov commited on

ci : disable failing CUDA and Java builds
ecef312
unverified

ggerganov commited on

readme : fix references to download-ggml-model.sh (#2427)
3d92452
unverified

Hugo commited on

make : remove "talk" target until updated
5fb8fce

ggerganov commited on

ggml : add ggml-cpu-impl.h (skip) (#0)
958f2d3

ggerganov commited on

sync : ggml
e22e2f8

ggerganov commited on

talk-llama : sync llama.cpp
f91f98d

ggerganov commited on

ggml : add AVX512DQ requirement for AVX512 builds (llama/9622)
14b5848

Eric Zhang commited on

log : add CONT level for continuing previous log entry (llama/9610)
a29a4c5

ggerganov commited on