ggml-org/llama.cpp

Pull Requests Commits

0cc4m committed 26 days ago

dba1cbce

0cc4m committed 26 days ago

94e2c4d2

0cc4m committed 26 days ago

c19b3c37

vulkan: allow MMQ bk_step tuning

0cc4m committed 27 days ago

7e8eb9ba

metal : remove obosolete asserts (#17295)

ggerganov committed 27 days ago

Verified 416e7c7f

server : handle context overflow during decode (#17267)

ggerganov committed 27 days ago

Verified 5b2093be

opencl: fix rms_norm_mul (#17250)

lhez committed 27 days ago

Verified 52e5d421

opencl: add kernel to handle mat mul in attention to improve encoding speed (#17181)

shaofeiqi committed 27 days ago

Verified 4db56412

sycl : unify unary kernels with a generic implementation and enable wide operator support (#17213)

shani-f committed 27 days ago

Verified 72bd7321

webui: Fix clickability around chat processing statistics UI (#17278)

allozaur committed 27 days ago

Verified 22e1ce2f

webui: add OAI-Compat Harmony tool-call streaming visualization and persistence in chat UI (#16618)

ServeurpersoCom committed 27 days ago

Verified 1411d927

convert : remove unnecessary chat template patching (#17289)

CISC committed 27 days ago

Verified 662192e1

vulkan: Fuse mul_mat_id+add_id+mul and mul_mat+add+add. (#17287)

jeffbolznv committed 27 days ago

Verified 24dc769f

vulkan: Replace 16-bit unpack8 calls to work around legacy Windows AMD driver bug (#17285)

0cc4m committed 28 days ago

Verified 4dca015b

convert : use all parts in safetensors index (#17286)

CISC committed 28 days ago

Verified 9a8860cf

convert : set expert gating func in base class (#17279)

CISC committed 28 days ago

Verified 9d3ef480

mtmd-cli: Avoid logging to stdout for model loading messages in mtmd-cli (#17277)

ankurvdev committed 28 days ago

Verified c7b7db04

vulkan: implement ABS and NEG (#17245)

giuseppe committed 28 days ago

Verified 1568d13c

vulkan: Use ggml_vk_tensor_subbuffer in mul_mat_vec(id) paths (#17244)

jeffbolznv committed 28 days ago

Verified 439342ea

vulkan: skip all-negative-inf blocks in FA (#17186)

jeffbolznv committed 28 days ago

Verified 234ae7d7

vulkan: change graph_compute to be async and enable get_tensor_async (#17158)

jeffbolznv committed 28 days ago

Verified 38eaf32a

mtmd: add mtmd_log_set (#17268)

ngxson committed 29 days ago

Verified 9b17d74a

model : add AfmoeForCausalLM support (#16477)

bartowski1182 committed 29 days ago

Verified e1fcf8b0

fix : Dangling pointer for non-empty trigger words in lazy grammar construction (#17048)

marek-hradil committed 29 days ago

Verified 6cd0cf72

server : fix "can batch with" bug (#17263)

ggerganov committed 29 days ago

Verified d396b437

metal : support argsort for ne00 > 1024 (#17247)

ggerganov committed 29 days ago

Verified 45c6ef73

metal : make the FA extra sizes consistent (#17143)

ggerganov committed 29 days ago

Verified 2606b0ad

readme : add RVV,ZVFH,ZFH,ZICBOP support for RISC-V (#17259)

ixgbe committed 29 days ago

Verified 307772fc

Better UX for handling multiple attachments in WebUI (#17246)

allozaur committed 29 days ago

Verified f1bad23f

ggml-cpu: handle 3d tensors in repack mat_mul (#17241)

Alcpz committed 29 days ago

Verified becc4816

Older