ggerganov/llama.cpp

Pull Requests Commits

server : do not speculate during prompt processing

ggerganov committed 173 days ago

Verified 33d7b70c

llama : add enum for built-in chat templates (#10623)

ngxson committed 173 days ago

Verified 642330ac

make : deprecate (#10514)

ggerganov committed 173 days ago

Verified 8648c521

server: Add "tokens per second" information in the backend (#10548)

lhpqaq committed 173 days ago

Verified 64ed2091

SYCL: Fix and switch to GGML_LOG system instead of fprintf (#10579)

qnixsynapse committed 174 days ago

Verified 991f8aab

contrib : refresh (#10593)

ggerganov committed 174 days ago

Verified 4cb003dd

Add `mistral-v1`, `mistral-v3`, `mistral-v3-tekken` and `mistral-v7` chat template types (#10572)

jukofyork committed 174 days ago

Verified 917786f4

grammars : add English-only grammar (#10612)

ggerganov committed 174 days ago

Verified 5e1ed955

ci: add error handling for Python venv creation in run.sh (#10608)

wangqin0 committed 174 days ago

Verified 5c7a5aa0

ggml : automatic selection of best CPU backend (#10606)

slaren committed 174 days ago

Verified 3420909d

server : bind to any port when specified (#10590)

alek3y committed 175 days ago

Verified 86dc11c5

readme : update the usage section with examples (#10596)

ggerganov committed 175 days ago

Verified 6acce397

build: update Makefile comments for C++ version change (#10598)

wangqin0 committed 175 days ago

Verified 43957ef2

ggml-cpu: replace AArch64 NEON assembly with intrinsics in ggml_gemv_q4_0_4x4_q8_0() (#10567)

angt committed 175 days ago

Verified 0c39f44d

readme : remove old badge

ggerganov committed 176 days ago

Verified 3e0ba0e6

readme : refresh (#10587)

ggerganov committed 176 days ago

Verified abadba05

vulkan: Dynamic subgroup size support for Q6_K mat_vec (#10536)

netrunnereve committed 176 days ago

Verified 0533e7fb

ggml : move AMX to the CPU backend (#10570)

slaren committed 176 days ago

Verified 7cc2d2c8

server : add more test cases (#10569)

ngxson committed 176 days ago

Verified b782e5c7

imatrix : support combine-only (#10492)

robbiemu committed 176 days ago

Verified 3a8e9af4

cleanup UI link list (#10577)

slaren committed 176 days ago

Verified a3a3048e

ggml : fix I8MM Q4_1 scaling factor conversion (#10562)

ggerganov committed 176 days ago

Verified f0678c5f

ggml-cpu: fix typo in gemv/gemm iq4_nl_4_4 (#10580)

FanShupei committed 176 days ago

Verified 4b3242bb

sycl : offload of get_rows set to 0 (#10432)

Alcpz committed 176 days ago

Verified 0f77aae5

sycl : Reroute permuted mul_mats through oneMKL (#10408)

Alcpz committed 177 days ago

Verified 266b8519

CANN: RoPE operator optimization (#10563)

noemotiovon committed 177 days ago

Verified 938f6087

vulkan: get the first command buffer submitted sooner (#10499)

jeffbolznv committed 177 days ago

Verified f095a649

llava: return false instead of exit (#10546)

tinglou committed 177 days ago

Verified 678d7994

ggml : remove redundant copyright notice + update authors

ggerganov committed 177 days ago

Verified dc223440

llama : add missing model types

ggerganov committed 177 days ago

Verified 4c0a95b1

Older