ngxson/llama.cpp

Pull Requests Commits

ui: npm run format for new upstream formatting rules

ServeurpersoCom committed 24 days ago

7c30fcd1

ui: store remote servers as a registry

ServeurpersoCom committed 24 days ago

254f9681

ui: derive the connecting state from the store

ServeurpersoCom committed 24 days ago

24998735

ui: move remote access constants and states out of the code

ServeurpersoCom committed 24 days ago

9bbe4d89

ui: show the splash while a restored session connects

ServeurpersoCom committed 24 days ago

69dcf755

ui: refresh backend state without reloading

ServeurpersoCom committed 24 days ago

47c538e0

add LLAMA_UI_WEB_ONLY option

ngxson committed 24 days ago

20bf1b04

ui: strip the base path from tunneled requests

ServeurpersoCom committed 24 days ago

ce1ca3e1

ui: fix remote access on reload and slow networks

ServeurpersoCom committed 24 days ago

fe84ab38

ui: announce the offer on all trackers at once

ServeurpersoCom committed 24 days ago

112fe96a

ui: replace the dead tracker in the signaling list

ServeurpersoCom committed 24 days ago

4e037c69

ui: fail loudly when the remote tunnel is down

ServeurpersoCom committed 24 days ago

e7806c22

ui: drop host mode, llama-connect handles it

ServeurpersoCom committed 24 days ago

739fd86c

ui: npm run format for new upstream formatting rules

ServeurpersoCom committed 24 days ago

acb0cd8a

ui: use arrow function in fetch interceptor

ServeurpersoCom committed 24 days ago

fba14fab

ui: (demo) access server remotely via webrtc

ngxson committed 24 days ago

9b73bf0f

sycl : enhance the api to support peer-to-peer copy (#27550)

arthw committed 24 days ago

Verified 0df974d7

convert : skip bias_vl tensor in DeepSeek-V4 DSpark conversion (#28294)

ggerganov committed 24 days ago

Verified d646c9d1

misc : prevent RAM peaking at model loading stage (#27483)

tdakhran committed 24 days ago

Verified 5ec4eab6

sycl: reduce redundant work in Q4_K multi-column MMVQ (#27062)

Eurekatic committed 24 days ago

Verified 4aa6ffba

model: add NVIDIA Nemotron-3-Puzzle-75B-A9B (NemotronHPuzzle) support (#25444)

YanissAmz committed 24 days ago

Verified c61b98b8

mtmd: fix idefics3 preproc (#28273)

ngxson committed 24 days ago

Verified 67a17c17

finetune: fix no KV cache (#27199)

ngxson committed 24 days ago

Verified 159b7414

server : accept data: URLs for input_video and input_audio (#27735)

geckguy committed 24 days ago

Verified 9cffdcc8

ggml-hexagon: add F16 support for unary ops (#28228)

cqderek committed 24 days ago

Verified f027c4f1

mtmd: add mtmd_tokenize_from_parts() (#28250)

ngxson committed 24 days ago

Verified 73390547

metal : add fa-vec tunings for M3 (#28236)

init-22 committed 24 days ago

Verified 9cc33944

metal : fix memory query under low-memory conditions (#27701)

madsmtm committed 24 days ago

Verified 8c0b9cd0

ci : check for missing autoreleasepools (#27884)

nikwen committed 24 days ago

Verified 03dbcc53

Update ROCm to 10.0.0 release (#27803)

superm1 committed 24 days ago

Verified cff18443

Older