Aldehir Rojas
|
c05aa69f32
common : add nemotron 3 parsing (#18077)
|
пре 1 месец |
Aldehir Rojas
|
2fbe3b7bb7
common : add parser for ministral/mistral large 3/devstral 2 (#17713)
|
пре 1 месец |
hksdpc255
|
636fc17a37
Fix Kimi-K2 tool-call parsing issues (#17376)
|
пре 1 месец |
hksdpc255
|
1920345c3b
common : Generalized XML-style tool-call parsing with streaming support (GLM 4.5/4.6 + MiniMax M2 + SeedOSS + Kimi-K2 + Qwen3-Coder + Apriel-1.5 + Xiaomi-MiMo) (#16932)
|
пре 1 месец |
Yuri Khrustalev
|
c053e18a66
chat: Add LFM2 tool handling (#16763)
|
пре 2 месеци |
Pascal
|
128d522c04
chat : support Magistral thinking (#16413)
|
пре 3 месеци |
Piotr Wilkin (ilintar)
|
34fcc5a4ac
model : Apertus model implementation (#15852)
|
пре 3 месеци |
shun095
|
f432d8d83e
chat: Fix streaming parser for granite models (#15682)
|
пре 3 месеци |
Jesse
|
88021565f0
chat : Deepseek V3.1 reasoning and tool calling support (OpenAI Style) (#15533)
|
пре 4 месеци |
Piotr Wilkin (ilintar)
|
b2426e469e
chat : nemotron thinking & toolcalling support (#15676)
|
пре 4 месеци |
Piotr Wilkin (ilintar)
|
60e5eee31f
chat : Seed OSS thinking + tool call support (#15552)
|
пре 4 месеци |
Xuan-Son Nguyen
|
e9288e8869
chat : clarify the meaning of reasoning_format (#15408)
|
пре 5 месеци |
Aldehir Rojas
|
b204a5a234
gpt-oss: implement harmony parsing (#15181)
|
пре 5 месеци |
Sachin Desai
|
3db4da56a5
chat : support Granite model reasoning and tool call (#14864)
|
пре 5 месеци |
Jhen-Jie Hong
|
f738989dcb
chat : fix multiple tool_calls on hermes-2-pro (#14962)
|
пре 5 месеци |
Diego Devesa
|
7f4fbe5183
llama : allow building all tests on windows when not using shared libs (#13980)
|
пре 7 месеци |
Olivier Chafik
|
c9bbc77931
`server`: update deepseek reasoning format (pass reasoning_content as diffs) (#13933)
|
пре 7 месеци |
Olivier Chafik
|
e15898d1c7
server: allow unclosed thinking tags (#13931)
|
пре 7 месеци |
Georgi Gerganov
|
53f925074d
sync : vendor (#13901)
|
пре 7 месеци |
Olivier Chafik
|
03f582ae8f
server: fix streaming crashes (#13786)
|
пре 7 месеци |
Olivier Chafik
|
d74e94c1b3
`server`: fix format of streamed tool call deltas (diff name, fix id location) (#13800)
|
пре 7 месеци |
Olivier Chafik
|
e121edc432
`server`: add `--reasoning-budget 0` to disable thinking (incl. qwen3 w/ enable_thinking:false) (#13771)
|
пре 7 месеци |
Olivier Chafik
|
f5cd27b71d
`server`: streaming of tool calls and thoughts when `--jinja` is on (#12379)
|
пре 7 месеци |
Olivier Chafik
|
aa48e373f2
`server`: inject date_string in llama 3.x template + fix date for firefunction v2 (#12802)
|
пре 8 месеци |
Georgi Gerganov
|
13b4548877
cmake : do not include ./src as public for libllama (#13062)
|
пре 8 месеци |
Olivier Chafik
|
b6930ebc42
`tool-call`: fix non-tool-calling grammar crashes w/ Qwen / Hermes 2 templates (#12900)
|
пре 9 месеци |
Olivier Chafik
|
4e39a3c332
`server`: extract <think> tags from qwq outputs (#12297)
|
пре 10 месеци |
Olivier Chafik
|
87c2630546
allow missing content in message if tool_calls provided (#12293)
|
пре 10 месеци |
Olivier Chafik
|
669912d9a5
`tool-call`: fix Qwen 2.5 Coder support, add micro benchmarks, support trigger patterns for lazy grammars (#12034)
|
пре 10 месеци |
Olivier Chafik
|
63e489c025
tool-call: refactor common chat / tool-call api (+ tests / fixes) (#11900)
|
пре 11 месеци |