ngxson/llama.cpp

Pull Requests Commits

decouple server_routes and server_http

ngxson committed 40 days ago

c5f8cdf3

fix windows build

ngxson committed 40 days ago

2141a3e7

also expose server_response_reader (to be used by CLI)

ngxson committed 40 days ago

8b510b47

ggerganov committed 40 days ago

Verified 26204aed

clean up headers

ngxson committed 40 days ago

239c7a26

add server-context.h

ngxson committed 40 days ago

9a7b4f36

add server-context.h

ngxson committed 40 days ago

0150602b

ngxson committed 40 days ago

3375c20b

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

arthw committed 40 days ago

Verified 7d2add51

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

ixgbe committed 40 days ago

Verified f698a79c

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

0cc4m committed 40 days ago

Verified 47a268ea

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

jeffbolznv committed 40 days ago

Verified 59d8d4e9

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

AlekseiNikiforovIBM committed 41 days ago

Verified d82b7a7c

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

dranger003 committed 41 days ago

Verified 03914c7e

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

o7si committed 41 days ago

Verified 3ce7a65c

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

slaren committed 41 days ago

Verified e072b205

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

yeahdongcn committed 41 days ago

Verified c6f7a423

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

am17an committed 41 days ago

Verified 2e7ef98f

server : add Anthropic Messages API support (#17570)

noname22 committed 41 days ago

Verified ddf9f943

model : Qwen3 Next (#16095)

pwilkin committed 41 days ago

Verified ff55414c

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

JohannesGaessler committed 41 days ago

Verified 73955f7d

vulkan: Implement GGML_OP_TRI (#17503)

jeffbolznv committed 41 days ago

Verified 35cf8887

rpc : cache and reuse compute graphs (#15405)

rgerganov committed 41 days ago

Verified 15d2b46b

HIP: enable mul_mat_f for RDNA4 (#17437)

zhang-hui-yulo committed 41 days ago

Verified 6bca76ff

SOLVE_TRI CUDA kernel for small matrices (#17457)

pwilkin committed 41 days ago

Verified cd0e3a7a

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

NeoZhangJianyu committed 41 days ago

Verified efaaccdd

vulkan: Implement SOLVE_TRI (#17486)

jeffbolznv committed 42 days ago

Verified 4abef75f

arch : add description about LLM_TENSOR_INFOS (#17550)

ggerganov committed 42 days ago

Verified c3861149

models : fix LFM2 tensors (#17548)

ggerganov committed 42 days ago

Verified 6783b11f

cuda : fix UMA detection on discrete GPUs. (#17537)

matt23654 committed 42 days ago

Verified 909072ab

Older