ggml-org/ggml

Pull Requests Commits

sync : whisper.cpp

ggerganov committed 1 year ago

Verified 3350042b

whisper : add check that target name exists (whisper/3103)

danbev committed 1 year ago

Verified 36edebc8

ggml : suppress Windows compiler warnings (whisper/3075)

danbev committed 1 year ago

Verified afe4322c

sync : llama.cpp

ggerganov committed 1 year ago

5d22a79e

CUDA: batched+noncont MMQ, refactor bs>1 MoE code (llama/13199)

JohannesGaessler committed 1 year ago

b5ed2968

vulkan: use uint array index to avoid glslang bug (llama/13193)

jeffbolznv committed 1 year ago

b2b4e0b0

ggml : fix ppc64le build (llama/13176)

shalinib-ibm committed 1 year ago

627cba58

feat(ggml-cpu): enable z17 compile (llama/13182)

taronaeo committed 1 year ago

656d93c3

CUDA: fix non-cont. inputs for batched mat mul (llama/13155)

JohannesGaessler committed 1 year ago

9fc15576

fix(rpc): Improve input validation and error handling (llama/13069)

thevilledev committed 1 year ago

92a571b4

SYCL: Add all missing unary kernels (llama/13074)

qnixsynapse committed 1 year ago

da1acd68

musa: fix typo in cc control (llama/13144)

yeahdongcn committed 1 year ago

16776a2e

CUDA: fix q_nope_absorbed prec for DS 2 Lite f16 (llama/13137)

JohannesGaessler committed 1 year ago

61fe29c8

musa: fix build warning (llama/13129)

yeahdongcn committed 1 year ago

c38c21cb

ggml: move fp16/bf16 conversion optimizations to CPU backend + export conversion APIs (llama/13107)

SongXiaoXi committed 1 year ago

ef1205f0

clip : fix pixtral on some GPU backends (llama/13097)

ngxson committed 1 year ago

5fe6d362

change the reorder tensor from init to execute OP (llama/13003)

NeoZhangJianyu committed 1 year ago

28aeed77

rpc : do not wait for response when sending RPC_CMD_SET_TENSOR (llama/12943)

rgerganov committed 1 year ago

c9207d32

ggml : fix ggml_gallocr_ptr type (#1205)

slaren committed 1 year ago

Verified 7734363a

media : rm logos (#1203)

ggerganov committed 1 year ago

Verified d50c1fff

sync : whisper.cpp

ggerganov committed 1 year ago

0eac9285

cuda : fix unused variable compile warning (whisper/0)

ggerganov committed 1 year ago

13d4d214

opencl : remove obsolete files (skip) (#1200)

ggerganov committed 1 year ago

Verified 489716ba

sync : llama.cpp

ggerganov committed 1 year ago

8d90ea0d

metal : add memory pool for temp allocs (llama/12850)

ggerganov committed 1 year ago

939718f7

opencl: split ggml-opencl.cl into multiple files and cleanup (llama/12886)

lhez committed 1 year ago

a7d9d656

ggml : fix trailing whitespaces (llama/0)

ggerganov committed 1 year ago

cc1f1245

CUDA: use switch statements in constexpr functions (llama/13095)

JohannesGaessler committed 1 year ago

c09add3c

metal : fix floating-point range of attention scores in FA kernels (llama/13090)

ggerganov committed 1 year ago

829c67c9

vulkan: matmul gcn tuning (llama/13016)

netrunnereve committed 1 year ago

5f7d009a

Older