ggml-org/llama.cpp

Pull Requests Commits

server : minor UI fix (#10207)

ngxson committed 314 days ago

Verified 76c6e7f1

server : revamp chat UI with vuejs and daisyui (#10175)

ngxson committed 314 days ago

Verified a71d81cf

scripts : add amx to sync-ggml.sh [no ci]

ggerganov committed 314 days ago

Verified eec4d717

ggerganov committed 314 days ago

Verified 3b088286

scripts : sync update

ggerganov committed 314 days ago

Verified a2c6fd74

ggml : add ggml-cpu.h to the public headers (#10204)

slaren committed 314 days ago

Verified 97404c4a

Remove identical wte/etw logic for jais (#10203)

fmz committed 314 days ago

Verified 60e17ce2

DRY: Fixes clone functionality (#10192)

wwoodsTM committed 314 days ago

Verified 5107e8ce

fix q4_0_8_8 format for corrupted tokens issue (#10198)

snadampal committed 315 days ago

Verified 2319126a

Optimize RWKV6 Operator Naming and Implement Multi-core CPU/ SYCL Acceleration (#10133)

zhiyuan1i committed 315 days ago

Verified 3bcd40b3

metal : add BF16 support (#8439)

ggerganov committed 315 days ago

Verified 5c333e01

server : remove hack for extra parallel slot (#10187)

ggerganov committed 315 days ago

Verified b11f9ba9

metal : fix from ptr buffer name (#10189)

slaren committed 315 days ago

Verified 94d8cb8b

ggml : adjust is_first_call init value (#10193)

ggerganov committed 316 days ago

Verified 1dc04b2d

metal : add quantized FA support (#10149)

ggerganov committed 316 days ago

Verified a1eaf6a9

llama : add <|tool_call|> formatting to Granite template (#10177)

gabe-l-hart committed 316 days ago

Verified b8deef0e

ggml : fix arch check in bf16_to_fp32 (#10164)

slaren committed 317 days ago

Verified a9e8a9a0

Q6_K AVX improvements (#10118)

netrunnereve committed 317 days ago

Verified 34073647

ggml : fix gelu tables initialization (#10172)

slaren committed 317 days ago

Verified d5a409e5

ggml : fix q4xx mat mul, increase ggml_aligned_malloc alignment (#10167)

slaren committed 317 days ago

Verified 401558b7

server : clarify /slots endpoint, add is_processing (#10162)

ngxson committed 317 days ago

Verified 9e0ecfb6

fix build break on arm64 linux (#10166)

snadampal committed 317 days ago

Verified 6a066b99

cuda : clear error after changing peer access (#10153)

slaren committed 317 days ago

Verified ea02c753

metal : simplify f16 and f32 dequant kernels (#0)

ggerganov committed 317 days ago

Verified 05697f67

metal : move dequantize templates to beginning of MSL source (#0)

ggerganov committed 317 days ago

Verified f8e58135

CANN: adjust backend registry refactor. (#10158)

leo-pony committed 317 days ago

Verified 329ed914

ggerganov committed 318 days ago

Verified ce027adf

cmake : make it possible linking ggml as external lib (ggml/1003)

ykhrustalev committed 318 days ago

Verified 284e5b02

metal : fix minor string leaks (ggml/1004)

pminev committed 318 days ago

Verified e2292aaa

ggml : move CPU backend to a separate file (#10144)

slaren committed 318 days ago

Verified 9f409893

Newer Older