ngxson/llama.cpp

Pull Requests Commits

use std::vector

ngxson committed 1 year ago

3fec232f

even better approach

ngxson committed 1 year ago

42a9c5a9

regenerate docs

ngxson committed 1 year ago

6b369601

common : improve ctv ctk cli argument

ngxson committed 1 year ago

ba35f29e

CUDA: faster non-contiguous concat (#10760)

A3shTnT committed 1 year ago

Verified 8faa1d4d

remove CMAKE_WINDOWS_EXPORT_ALL_SYMBOLS (#10797)

slaren committed 1 year ago

Verified cb13ef85

Vulkan: Use improved q4_k and q5_k dequant code in dequant shaders (#10798)

0cc4m committed 1 year ago

Verified 4064c0e3

Vulkan: Add VK_EXT_subgroup_size_control support to ensure full subgroups for coopmats (#10721)

0cc4m committed 1 year ago

Verified dc5301d5

common : add missing env var for speculative (#10801)

ngxson committed 1 year ago

Verified 9fdb1243

docs: update server streaming mode documentation (#9519)

CentricStorm committed 1 year ago

Verified 5555c0c1

Merge pull request #10788 from ggerganov/gg/gguf-py-0.11.0

ggerganov committed 1 year ago

Verified 973f328b

gguf-py : bump version to 0.11.0

ggerganov committed 1 year ago

Verified fb18934a

server : (UI) add tok/s, get rid of completion.js (#10786)

ngxson committed 1 year ago

Verified 235f6e14

Update README.md (#10772)

qingy1337 committed 1 year ago

Verified 1a31d0dc

ci : pin nodejs to 22.11.0 (#10779)

ngxson committed 1 year ago

Verified 92f77a64

bug-fix: snprintf prints NULL in place of the last character (#10419)

kallewoof committed 1 year ago

Verified 484d2f31

docs: fix server documentation formatting (#10776)

CentricStorm committed 1 year ago

Verified 4b4d92b0

ggml: load all backends from a user-provided search path (#10699)

giladgd committed 1 year ago

Verified 43041d2e

vulkan: request round-to-even for fp16 in im2col/rope_head (#10767)

jeffbolznv committed 1 year ago

Verified b685daf3

vulkan: dynamic subgroup size for the remaining k quants (#10745)

netrunnereve committed 1 year ago

Verified dafae66c

imatrix : Add imatrix to --no-context-shift (#10766)

bartowski1182 committed 1 year ago

Verified ae4b9226

CUDA: rename macros to avoid conflicts with WinAPI (#10736)

aendk committed 1 year ago

Verified 750cb3e2

server : add flag to disable the web-ui (#10762) (#10751)

eugeniosegala committed 1 year ago

Verified a86ad841

vulkan: disable spirv-opt for coopmat shaders (#10763)

jeffbolznv committed 1 year ago

Verified a05e2afc

CUDA: fix shared memory access condition for mmv (#10740)

JohannesGaessler committed 1 year ago

Verified 26a8406b

Changes to CMakePresets.json to add ninja clang target on windows (#10668)

Srihari-mcw committed 1 year ago

Verified c37fb4cf

vulkan: fix compile warnings (#10731)

jeffbolznv committed 1 year ago

Verified 3d98b4cb

cmake : simplify msvc charsets (#10672)

iboB committed 1 year ago

Verified 1a050047

server : fix format_infill (#10724)

ngxson committed 1 year ago

Verified ce8784bd

server : bring back info of final chunk in stream mode (#10722)

ngxson committed 1 year ago

Verified e52522b8

Older