Follow
The Wire

Alibaba, newest first.

Linked and quoted from the labs, the press and the people using them. Times in UTC.

The Wire

Top storiesEverything

Every development we picked up, including smaller stories beyond the labs. Newest first, times in UTC.

AllLaunchesUpdatesIncidentsResearchPolicy and money

Thursday 24 September

1 development
08:36
AlibabaUpdate

llama.cpp adds support for Ling 3.0 VL model

A pull request adds Ling 3.0 VL support to llama.cpp. Ling-3.0-flash-VL has 124B total parameters with 5.5B activated per token and a 256K token context window.

1 community

Tuesday 22 September

1 development
18:31
AlibabaUpdate

Qwen Image 2.1 FP8 runs under 10GB in Unsloth Studio

A community post shows Qwen Image 2.1 in a Fast FP8 build generating high quality images. The user runs it on Unsloth Studio and notes an update is needed to see the option.

1 community
Older

Alibaba on the Score