Skip to content

Get up and running with Kimi, GLM, MiniMax, DeepSeek, gpt-oss, Qwen, Gemma and other models.

Commits on main

ggml: Backport scale kernel fixes
Jesse Gross committed 1 year ago
efaee8c
ggml: Remove allocation status reporting
Jesse Gross committed 1 year ago
734b57d
qwen3-coder: fix tool definition type rendering
Devon Rifkin committed 1 year ago
83021fc
0469861
parsers: fix unicode handling for qwen3-coder
Devon Rifkin committed 1 year ago
05ba4ca
5a56ff3
Grace/deepseek v3 migration (#12385)
Grace committed 1 year ago
fbd82ba
prefer ollama engine for qwen3moe (#12374)
Michael Yang committed 1 year ago
2e74254
bbb195a
harmony: don't sanitize built-ins
Devon Rifkin committed 1 year ago
fd88cd7
fix: leaf alt name (#12390)
Michael Yang committed 1 year ago
e1979c5
add pre:, suf: to tags (#12274)
Michael Yang committed 1 year ago
bf78ed6
multi-regexp pretokenizer (#12325)
Michael Yang committed 1 year ago
a40d427
auth: fix problems with the ollama keypairs (#12373)
Patrick Devine committed 1 year ago
64883e3
tests: add single threaded history test (#12295)
Daniel Hiltgen committed 1 year ago
c23e6f4
docs: update cloud.md for cloud models
jmorganca committed 1 year ago
af060eb
docs: move turbo.md to cloud.md
jmorganca committed 1 year ago
ae5c330
parsers: fix `&`s in qwen3coder parameter values
Devon Rifkin committed 1 year ago
242df70
gemma: fix rope scaling for qat models (#12348)
Patrick Devine committed 1 year ago
dba39b2
9f3a37f
feat: qwen3 embed (#12301)
Michael Yang committed 1 year ago
7460259
22ccdd7
build: avoid unbounded parallel builds (#12319)
Daniel Hiltgen committed 1 year ago
0c3d0e7
harmony: remove special casing in routes.go
Devon Rifkin committed 1 year ago
e7f56ef