ngxson/llama.cpp

Pull Requests Commits

cli: add migration warning

ngxson committed 19 days ago

3632271e

common: add LLAMA_LOG_FILE env var (#17609)

taronaeo committed 19 days ago

Verified def5404f

ggml: fix: macOS build with `-DGGML_BACKEND_DL=ON` (#17581)

giladgd committed 19 days ago

Verified fa046595

common: update env var name (#17588)

ddh0 committed 19 days ago

Verified 5a6241fe

CUDA: add stream-based concurrency (#16991)

am17an committed 19 days ago

Verified c7af376c

cuda : add error checking for cudaMemcpyAsync in argsort (#17599)

Mahekk357 committed 19 days ago

Verified 00425e2e

vulkan : fix FA mask load with bounds check (coopmat2) (#17606)

Acly committed 19 days ago

Verified 385c3da5

server: move server-context to its own cpp|h (#17595)

ngxson committed 20 days ago

Verified ab49f094

server: explicitly set the function name in lambda (#17538)

haiyuewa committed 20 days ago

Verified 8c32d9d9

common : fix json schema with '\' in literals (#17307)

i-v-s committed 20 days ago

Verified 0874693b

sycl : support to malloc memory on device more than 4GB, update the doc and script (#17566)

arthw committed 20 days ago

Verified 7d2add51

ggml: replace hwcap with riscv_hwprobe for RVV detection (#17567)

ixgbe committed 20 days ago

Verified f698a79c

Vulkan: MMVQ Integer Dot K-Quant and MUL_MAT_ID support (#16900)

0cc4m committed 20 days ago

Verified 47a268ea

vulkan: improve topk perf for large k, fix overflow in unit tests (#17582)

jeffbolznv committed 20 days ago

Verified 59d8d4e9

gguf-py : fix passing non-native endian tensors (editor-gui and new-metadata) (#17553)

AlekseiNikiforovIBM committed 21 days ago

Verified d82b7a7c

common : move all common_chat_parse_* to chat-parser.cpp. (#17481)

dranger003 committed 21 days ago

Verified 03914c7e

server: fix: /metrics endpoint returning JSON-escaped Prometheus format (#17386)

o7si committed 21 days ago

Verified 3ce7a65c

ggml : add GGML_SCHED_NO_REALLOC option to disable reallocations in ggml_backend_sched (#17276)

slaren committed 21 days ago

Verified e072b205

[MUSA] enable fp16/fast_fp16/bf16_mma on PH1 (#17551)

yeahdongcn committed 21 days ago

Verified c6f7a423

ggml-cuda: add stricter checking for fusion (#17568)

am17an committed 21 days ago

Verified 2e7ef98f

server : add Anthropic Messages API support (#17570)

noname22 committed 21 days ago

Verified ddf9f943

model : Qwen3 Next (#16095)

pwilkin committed 21 days ago

Verified ff55414c

CUDA: no FP16 arithmetic for vector FA kernel (#17558)

JohannesGaessler committed 21 days ago

Verified 73955f7d

vulkan: Implement GGML_OP_TRI (#17503)

jeffbolznv committed 21 days ago

Verified 35cf8887

rpc : cache and reuse compute graphs (#15405)

rgerganov committed 21 days ago

Verified 15d2b46b

HIP: enable mul_mat_f for RDNA4 (#17437)

zhang-hui-yulo committed 21 days ago

Verified 6bca76ff

SOLVE_TRI CUDA kernel for small matrices (#17457)

pwilkin committed 21 days ago

Verified cd0e3a7a

refactor pad_reflect_1d to make the UT case pass (#17204)

NeoZhangJianyu committed 21 days ago

Verified efaaccdd

vulkan: Implement SOLVE_TRI (#17486)

jeffbolznv committed 22 days ago

Verified 4abef75f

arch : add description about LLM_TENSOR_INFOS (#17550)

ggerganov committed 22 days ago

Verified c3861149

Older