ggerganov/llama.cpp

Pull Requests Commits

llama : support GEGLU for jina-bert-v2 (#14090)

CISC committed 193 days ago

Verified 3678b838

vulkan: force device 0 in CI (#14106)

jeffbolznv committed 193 days ago

Verified 652b70e6

Fixed spec timings to: accepted/tested instead of accepted/drafted (#14104)

jukofyork committed 193 days ago

Verified 3a12db23

ggerganov committed 193 days ago

ae92c185

ggml : fix weak alias win32 (whisper/0)

ggerganov committed 193 days ago

b7ce1ad1

Vulkan: Don't default to CPU device (like llvmpipe), even if no other device is available, to allow fallback to CPU backend (#14099)

0cc4m committed 194 days ago

Verified 97340b4c

rpc : nicer error messages for RPC server crash (#14076)

isaac-mcfadyen committed 194 days ago

Verified 2bb04670

ggerganov committed 194 days ago

b8e2194e

Add in-build ggml::ggml ALIAS library (ggml/1260)

dg0yt committed 194 days ago

1a3b5e80

metal : use less stack memory in FA kernel (#14088)

ggerganov committed 194 days ago

Verified 1f63e75f

kv-cache : fix shift and defrag logic (#14081)

ggerganov committed 194 days ago

Verified 40cbf571

llama : allow building all tests on windows when not using shared libs (#13980)

slaren committed 194 days ago

Verified 7f4fbe51

ggml-cpu : split arch-specific implementations (#13892)

xctan committed 194 days ago

Verified f470bc36

cuda : fix device sync on buffer clear (#14033)

slaren committed 194 days ago

Verified 8f47e25f

graph : fix geglu (#14077)

ggerganov committed 194 days ago

Verified 201b31dc

CANN: Simplify the environment variable setting(#13104)

bachelor-dou committed 195 days ago

Verified e21d2d4a

webui: fix sidebar being covered by main content (#14082)

yeahdongcn committed 195 days ago

Verified dc0623fd

server : fix LRU check (#14079)

ggerganov committed 195 days ago

Verified 87d34b38

sycl: Add reorder to Q6_K mmvq implementation (#13885)

s-Nick committed 195 days ago

Verified b460d16a

add geglu activation function (#14074)

huydt84 committed 195 days ago

Verified 91a8ee6a

CANN: Enable labeler for Ascend NPU (#13914)

shink committed 195 days ago

Verified 056eb745

cuda : fix buffer type check with integrated GPUs (#14069)

slaren committed 195 days ago

Verified 247e5c6e

ci: add LoongArch cross-compile build (#13944)

wojiushixiaobai committed 196 days ago

Verified 5787b5da

SYCL: Implement few same quantized type copy kernels (#13739)

qnixsynapse committed 197 days ago

Verified 228f34c9

llama : fix llama_model_chat_template with template name (LLM_KV with suffix) (#14050)

CISC committed 197 days ago

Verified 0974ad7a

llama : deprecate llama_kv_self_ API (#14030)

ggerganov committed 198 days ago

Verified 745aa531

context : fix SWA-related warning for multiple sequences (#14045)

ggerganov committed 198 days ago

Verified 487a5e04

llama : support multiple classifier outputs and labels (#13940)

CISC committed 198 days ago

Verified d17a809e

gguf-py : add add_classifier_output_labels method to writer (#14031)

CISC committed 198 days ago

Verified 1caae7fc

vulkan: Enable VK_KHR_cooperative_matrix extension for Intel Xe2 GPUs (#14001)

rillomas committed 198 days ago

Verified 669c13e0

Newer Older