ggerganov/llama.cpp

Pull Requests Commits

ggerganov committed 102 days ago

Verified a5801f40

cmake : Indent ggml-config.cmake (ggml/1310)

dg0yt committed 102 days ago

Verified 2c1f8101

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

ORippler committed 102 days ago

Verified e2b7621e

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

ggerganov committed 102 days ago

Verified c1dbea75

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

kiwi142857 committed 102 days ago

Verified 749e0d27

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

struct committed 102 days ago

Verified 64bf1c37

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

slaren committed 102 days ago

Verified c12bbde3

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

yeahdongcn committed 103 days ago

Verified 3f4fc97f

ggerganov committed 103 days ago

2df255da

cmake : fix usage issues (ggml/1257)

dg0yt committed 103 days ago

60f816a7

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

danbev committed 103 days ago

5592f278

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

ggerganov committed 103 days ago

Verified e4868d16

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

ngxson committed 103 days ago

Verified 820de57d

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

Alcpz committed 103 days ago

Verified cb4a63aa

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

jk3456a committed 103 days ago

Verified 86f5623d

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

PouyaGhahramanian committed 103 days ago

Verified 39cffdf1

metal : fix fusion across different encoders (#14849)

ggerganov committed 103 days ago

Verified 065908cb

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

djeong20 committed 104 days ago

Verified 4ec6291a

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

jacekpoplawski committed 104 days ago

Verified a12363bb

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

JohannesGaessler committed 104 days ago

Verified a86f52b2

CUDA: fix compilation with GGML_CUDA_F16 (#14837)

JohannesGaessler committed 104 days ago

Verified b284197d

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

CISC committed 104 days ago

Verified 221c0e0c

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

JohannesGaessler committed 104 days ago

07a19e27

tests : add non-cont K,V FA tests

ggerganov committed 104 days ago

18f3b5ff

memory : handle saving/loading null layers in recurrent memory (#14675)

l3utterfly committed 104 days ago

Verified 7233358d

ggml: fix loongarch quantize_row_q8_1 error (#14827)

lixing-star committed 104 days ago

Verified 6c88b3bb

CANN: weight format to NZ for Ascend310P3 (#14407)

tqgy6 committed 105 days ago

Verified 14c28dfc

CUDA: add fused rms norm (#14800)

am17an committed 105 days ago

Verified 8c988fa4

ggml : model card yaml tab->2xspace (#14819)

csabakecskemeti committed 105 days ago

Verified acd6cb1c

vulkan: fix rms_norm_mul to handle broadcasting dim0 (#14817)

jeffbolznv committed 105 days ago

Verified 84712b60

Older