ggerganov/llama.cpp

Pull Requests Commits

ggerganov committed 1 year ago

381879e0

repack : optimize mul_mat_id path

ggerganov committed 1 year ago

Verified 477d4398

ggml : repack block_iq4_nlx8

ggerganov committed 1 year ago

e2661edd

CANN: Implement GLU ops (#14884)

hipudding committed 1 year ago

Verified 11dd5a44

musa: fix build warnings (unused variable) (#14869)

yeahdongcn committed 1 year ago

Verified 9b8f3c6c

ggml-cpu : disable GGML_NNPA by default due to instability (#14880)

taronaeo committed 1 year ago

Verified c7f3169c

metal: SSM_SCAN performance (#14743)

gabe-l-hart committed 1 year ago

Verified 793c0d7f

opencl: add fused `rms_norm_mul` (#14841)

lhez committed 1 year ago

Verified ce111d39

docs : update HOWTO‑add‑model.md for ModelBase and new model classes (#14874)

wooksong committed 1 year ago

Verified e7fecba9

ggml : remove invalid portPos specifiers from dot files (#14838)

ORippler committed 1 year ago

Verified e2b7621e

context : restore preemptive sched reset when LLAMA_SET_ROWS=0 (#14870)

ggerganov committed 1 year ago

Verified c1dbea75

mtmd : fix 32-bit narrowing issue in export-lora and mtmd clip (#14503)

kiwi142857 committed 1 year ago

Verified 749e0d27

rpc : check for null buffers in get/set/copy tensor endpoints (#14868)

struct committed 1 year ago

Verified 64bf1c37

sched : fix multiple evaluations of the same graph with pipeline parallelism (#14855)

slaren committed 1 year ago

Verified c12bbde3

musa: upgrade musa sdk to rc4.2.0 (#14498)

yeahdongcn committed 1 year ago

Verified 3f4fc97f

ggerganov committed 1 year ago

2df255da

cmake : fix usage issues (ggml/1257)

dg0yt committed 1 year ago

60f816a7

ggml-cpu : remove stdlib include from repack.cpp (ggml/1276)

danbev committed 1 year ago

5592f278

context : perform output reorder lazily upon access after sync (#14853)

ggerganov committed 1 year ago

Verified e4868d16

chat : fix kimi-k2 chat template (#14852)

ngxson committed 1 year ago

Verified 820de57d

sycl: fixed semantics of block offset calculation (#14814)

Alberto Cabrera Pérez committed 1 year ago

Verified cb4a63aa

llama : fix MiniCPM inference after Granite Four changes (#14850)

jk3456a committed 1 year ago

Verified 86f5623d

docs: add libcurl-dev install hint for Linux distros (#14801)

PouyaGhahramanian committed 1 year ago

Verified 39cffdf1

metal : fix fusion across different encoders (#14849)

ggerganov committed 1 year ago

Verified 065908cb

sycl: fix undefined variable in work group size check (#14843)

djeong20 committed 1 year ago

Verified 4ec6291a

convert : text-only support for GLM-4.1V-9B-Thinking (#14823)

jacekpoplawski committed 1 year ago

Verified a12363bb

CUDA: fix overflow in FA, tune performance (#14840)

JohannesGaessler committed 1 year ago

Verified a86f52b2

CUDA: fix compilation with GGML_CUDA_F16 (#14837)

JohannesGaessler committed 1 year ago

Verified b284197d

ci : correct label refactor->refactoring (#14832)

CISC committed 1 year ago

Verified 221c0e0c

CUDA: fix quantized KV cache + multiple sequences (#14822)

JohannesGaessler committed 1 year ago

07a19e27

Older