ggerganov/llama.cpp

Pull Requests Commits

server : fix typo in comment

compilade committed 283 days ago

93ec58b9

Merge branch 'master' into compilade/fix-server-long-system-prompt

compilade committed 283 days ago

af2f84c9

server : fix parallel generation with very small batch sizes

compilade committed 283 days ago

c1b738ef

ggml : move rope type enum to ggml.h (#8949)

danbev committed 283 days ago

Verified 06943a69

export-lora : throw error if lora is quantized (#9002)

ngxson committed 283 days ago

Verified 828d6ff7

ci : fix github workflow vulnerable to script injection (#9008)

diogoteles08 committed 284 days ago

Verified fc4ca27b

ci : enable RPC in all of the released builds (#9006)

rgerganov committed 284 days ago

Verified 1f67436c

llama : model-based max number of graph nodes calculation (#8970)

nicoboss committed 284 days ago

Verified 0fd93cde

docs: introduce gpustack and gguf-parser (#8873)

thxCode committed 284 days ago

Verified 84eb2f4f

grammar-parser : fix possible null-deref (#9004)

DavidKorczynski committed 284 days ago

Verified 1262e7ed

ggml: fix div-by-zero (#9003)

DavidKorczynski committed 284 days ago

Verified df5478fb

Fix a spelling mistake (#9001)

Septa2112 committed 284 days ago

Verified 2589292c

py : fix requirements check '==' -> '~=' (#8982)

ggerganov committed 284 days ago

Verified d3ae0ee8

server : handle models with missing EOS token (#8997)

ggerganov committed 285 days ago

Verified 5ef07e25

gguf-py : Numpy dequantization for most types (#8939)

compilade committed 285 days ago

Verified 4134999e

server : fix segfault on long system prompt

compilade committed 285 days ago

7eda5583

flake.lock: Update (#8979)

ggerganov committed 285 days ago

Verified 8cd1bcfd

update guide (#8909)

arthw committed 285 days ago

Verified a21c6fd4

llama : check all graph nodes when searching for result_embd_pooled (#8956)

fairydreaming committed 285 days ago

Verified 33309f66

Optimize Vulkan backend for better CPU performance and less GPU synchronization overhead. (#8943)

mtavenrath committed 285 days ago

Verified 7c5bfd57

metal : fix uninitialized abort_callback (#8968)

slaren committed 286 days ago

Verified 6e02327e

llama : default n_swa for phi-3 (#8931)

ngxson committed 286 days ago

Verified 7eb23840

Add support for encoder-only T5 models (#8900)

fairydreaming committed 286 days ago

Verified 7c3f55c1

gguf-py : fix double call to add_architecture() (#8952)

tarilabs committed 287 days ago

Verified 911b437f

Merge commit from fork

ggerganov committed 287 days ago

Verified b72942fa

llama : add support for lora adapters in T5 model (#8938)

fairydreaming committed 287 days ago

Verified 6afd1a99

make : fix llava obj file race (#8946)

ggerganov committed 287 days ago

Verified 272e3bd9

llama : better replace_all (cont) (#8926)

ggerganov committed 287 days ago

Verified 45a55b91

llava : support MiniCPM-V-2.5 (#7599)

tc-mb committed 287 days ago

Verified 3071c0a5

ggerganov committed 288 days ago

Verified 4305b57c

Older