ochafik/llama.cpp

Pull Requests Commits

llama 3.1: allow `{name:` & `{function:` syntax even w/ builtin tools (70B model just likes that!)

ochafik committed 1 year ago

774557cf

Unify content + message in server_task_result_cmpl_final (+ avoid string copy)

ochafik committed 1 year ago

d86a1ae8

Avoid passing tools twice in generic handler (now that minja passes them automatically when needed)

ochafik committed 1 year ago

77c60e66

Partial revert of LLAMA_CACHE=tmp (unless set explicitly in env)

ochafik committed 1 year ago

a810c37c

Add tool call to hot topics

ochafik committed 1 year ago

cbecb356

Somehow /* bad inside block comments, ok fine.

ochafik committed 1 year ago

64545ac9

Add cli mode to test-chat to generate template summaries markdown

ochafik committed 1 year ago

2b245697

Remove server tests LLAMA_CACHE override (tests are serial, and the cache is easier to prefill w/ scripts/fetch_server_test_models.py)

ochafik committed 1 year ago

84bc083f

ochafik committed 1 year ago

bc8a6113

Finish renaming of chat inputs vs. params [skip ci]

ochafik committed 1 year ago

36c776f3

Rename: common/chat.*, common_chat_{inputs -> params}

ochafik committed 1 year ago

ed7c622d

ochafik committed 1 year ago

6e676c80

Unify llama 3.x chat handling again (allow `{"type": "function", "name": ...` prefix)

ochafik committed 1 year ago

ba27e985

server : add /apply-template endpoint for additional use cases of Minja functionality (#11489)

pnb committed 1 year ago

Verified eb7cf15a

Move templates/ under models/

ochafik committed 1 year ago

7b5e0803

Create meta-llama-Llama-3.1-8B-Instruct.jinja

ochafik committed 1 year ago

682026f8

Merge remote-tracking branch 'origin/master' into tool-call

ochafik committed 1 year ago

babdefc4

ochafik committed 1 year ago

0f8af536

tool-calls: disable crashing tests

ochafik committed 1 year ago

77dd67c2

vulkan: implement initial support for IQ2 and IQ3 quantizations (#11360)

remyoudompheng committed 1 year ago

Verified 66ee4f29

Update test_tool_call.py

ochafik committed 1 year ago

76f6ab19

rm unused templates, rename one

ochafik committed 1 year ago

41eec462

Merge branch 'tool-call' of github.com:ochafik/llama.cpp into tool-call

ochafik committed 1 year ago

40cc3f2f

Split bulk of tool call tests to slow lane

Olivier Chafik committed 1 year ago

384f54a1

rm dead code + nits

Olivier Chafik committed 1 year ago

923c805d

server : update auto gen files comments [no ci] (#11484)

danbev committed 1 year ago

Verified e51c47b4

vulkan: Catch pipeline creation failure and print an error message (#11436)

jeffbolznv committed 1 year ago

Verified 2711d021

Parse https://ollama.com/library/ syntax (#11480)

ericcurtin committed 1 year ago

Verified f0d4b29e

ggerganov committed 1 year ago

Verified 81585779

ggml : add option to not print stack on abort (ggml/1081)

WilliamTambellini committed 1 year ago

Verified 1a0e87d2

Newer Older