Open Source LLM

Alibaba AI model-based agentic framework tops global ranking

Artificial intelligence agents are software programs that are capable of autonomously performing tasks on behalf of a user or another system. Photo: Shutterstock

Jointly developed by open-source initiative Agentica and San Francisco-based start-up Together AI, DeepSWE was trained on the Qwen3-32B large language model (LLM) – part of Alibaba Cloud’s third-generation family of AI models. It topped the leaderboard of the latest SWEBench-Verified test, scoring 59 per cent accuracy against other so-called open-weight models like DeepSeek ’s V3-0
Continue reading...

Open Source LLM

ZLUDA Making Progress In 2025 On Bringing CUDA To Non-NVIDIA GPUs

The ZLUDA open-source effort that started off a half-decade ago as a drop-in CUDA implementation for Intel GPUs and then for several years was funded by AMD as a CUDA implementation for Radeon GPUs atop ROCm and then open-sourced but then reverted has been continuing to push along a new path since last year. The current take on ZLUDA is a multi-vendor CUDA implementation for non-NVIDIA GPUs for AI workloads and more. More progress was made during Q2 on this effort.

The Q2
Continue reading...

Open Source LLM

Scientist Use A.I. To Mimic the Mind, Warts and All

To better understand human cognition, scientists trained a large language model on 10 million psychology experiment questions. It now answers questions much like we do.

Open Source LLM

ByteDance’s Douyin launches Yelp-like AI chatbot to suggest eateries, reviews

Tanfan, which translates to “discover meals” in English, is only accessible via the Chinese short video app Douyin. Photo: Shutterstock

The chatbot called Tanfan, which translates to “discover meals” in English, is only accessible via the Chinese short video app Douyin, where users are also able to book restaurants and order dishes from certain eateries.

As of Wednesday, Tanfan had registered about 35,000 users, according to ByteDance data. Besides providing text information, Tanfan can also be used to compare selected restaurants and watch food video blogs.

Powered by ByteDance’
Continue reading...

Open Source LLM

中國自研編程語言 華為「倉頡」開放下載 月底將正式開源

華為自主研發的倉頡編程語言正式推出首個長期支援(Long-Term Support,LTS)版本 1.0.0,現已開放 Windows、Mac 及 Linux 版本的 SDK 下載( )。該公司表示,LTS 版本專為企業級開發及大型工程項目而設,每兩年發布一次並提供長達三年的官方

Continue reading...

Open Source LLM

Huawei to open-source self-developed programming language Cangjie

The logo of Huawei is seen at the Viva Technology conference in Paris, France, in June. Photo: Reuters

Huawei Technologies is set to open-source its self-developed programming language, Cangjie, marking the latest step in the company’s pursuit of technological self-sufficiency.

First unveiled a year ago, Cangjie will be open-sourced and accessible to all developers starting July 30, Huawei announced during its annual developer conference earlier this month.

Open-sourcing allows public access to a software program’s source code, enabling third-party developers

Continue reading...

Open Source LLM

Exclusive: Grammarly acquires email startup Superhuman in AI platform push - Reuters

Exclusive: Grammarly acquires email startup Superhuman in AI platform push Reuters

Open Source LLM

digiKam 8.7 Released With AI Auto-Rotation Tool, OpenCV OpenCL + CUDA Support

The digiKam 8.7 professional photo management software has been released for this KDE/Qt-aligned open-source solution for digital photography.

Similar to prior digiKam releases working on new AI features , that AI work has continued with digiKam 8.7. New to digiKam 8.7 is having an AI auto-rotation tool to automate the orientation detection through content analysis using deep learning. This plug-in can then auto-rotate images when deemed necessary.

Continue reading...

Open Source LLM

Baidu joins open-source movement by making Ernie 4.5 models publicly available

Chinese tech giant Baidu on Monday marked its entry into the highly competitive field of Chinese open-source artificial intelligence (AI) systems, by making its flagship Ernie 4.5 models available for download on AI site Hugging Face.

Baidu open-sourced 10 variants from its Ernie 4.5 multimodal model family, from the 0.3 billion parameter lightweight models to the heavyweight 424 billion parameter ones, according to a statement.

Beijing-based Baidu, one of the earliest tech firms in China to develop

The logo of Baidu’s Ernie Bot is displayed near a screen showing the Baidu logo, in this illustration picture taken June 28, 2023. Photo: Reuters

Open Source LLM

百度正式開源文心 4.5 系列模型 最大模型總參數量達 424B

AI 模型開源化已成趨勢,百度最近就終於正式開源文心 4.5 系列模型,一次推出 10 款模型,涵蓋 47B、3B 參數的混合專家(MoE)模型,以及 0.3B 參數的稠密型模型。

文心 4.5 開源系列針對 MoE 架構提出創新性的多模

Continue reading...