Commits

History for

Commits on Jul 21, 2026

Fixes in response to copilot review, such as erroring with a 400 instead of a 500 on the wrong json type and handling cultures that use commas in numbers.

simonsoanes
authored and
zhongkaifu
committed
bbc89b0
Replaced the concurrent dictionary with a memory cache so we don't exhaust available memory.

simonsoanes
authored and
zhongkaifu
committed
a76fb65
Added initial support for the OpenAI Responses API and a basic in-memory response store that doesn't support replication.

simonsoanes
authored and
zhongkaifu
committed
350fc20

Commits on Jul 20, 2026

fix super long context loading issue
zhongkaifu
committed
47e6f67

Commits on Jul 18, 2026

improve cuda backend performance
zhongkaifu
committed
0fb0da7

Commits on Jul 13, 2026

Remove native debug/timing paths, redundant CPU A/B kill-switches, housekeeping

zhongkaifu
and
claude
committed
4175be0

Commits on Jul 12, 2026

support multi-images edit
zhongkaifu
committed
7c5e15d

Commits on Jul 10, 2026

update documents
zhongkaifu
committed
fd684cc

Commits on Jul 9, 2026

show temporary edited images
zhongkaifu
committed
0789c8d

Commits on Jul 7, 2026

support PDF format document
zhongkaifu
committed
509d1e1

Commits on Jul 6, 2026

update documents and fix broken builds
zhongkaifu
committed
09b48b9
support vulkan backend
zhongkaifu
committed
1709d1c

Commits on Jul 2, 2026

optimize image edit
zhongkaifu
committed
e30b52b

Commits on Jul 1, 2026

prefill optimization
zhongkaifu
committed
4631a25
update
zhongkaifu
committed
65dd2dd
optimize image edit
zhongkaifu
committed
84480be
optimization image edit
zhongkaifu
committed
879b85b

Commits on Jun 29, 2026

fix parallel requests bugs in Qwen
zhongkaifu
committed
5f2f60d

Commits on Jun 28, 2026

update wikipedia
zhongkaifu
committed
7e2bafc
optimization batch inference
zhongkaifu
committed
b9ab163
optimize
zhongkaifu
committed
7251e88
support Qwen Image Edit model
zhongkaifu
committed
cf5de09

Commits on Jun 23, 2026

optimize Qwen3.6
zhongkaifu
committed
ab4e2a7
optimize Qwen3.6
zhongkaifu
committed
1a091b8

Commits on Jun 17, 2026

improve cuda backend performance on gemma-4 mtp model
zhongkaifu
committed
99f1d7d
support gemma4 mtp
zhongkaifu
committed
ab2f4fc
support Qwen3.6 mtp
zhongkaifu
committed
1ac7153

Commits on Jun 12, 2026

support diffusion gemma model
zhongkaifu
committed
23f7a65

Commits on Jun 11, 2026

update dockerfile gpu
zhongkaifu
committed
99d6d83

Commits on Jun 10, 2026

update web chat ux
zhongkaifu
committed
8c7fc17
adjust server url
zhongkaifu
committed
6506a4d
add request processing metrics
zhongkaifu
committed
e803b82
Add dockerfile
zhongkaifu
committed
5d4ccc3

Commits on Jun 8, 2026

fix gemma ua load issue
zhongkaifu
committed
0af2a50
support gemma 4 uv
zhongkaifu
committed
a0a9abf