Search AI Products and News

Recently, Alibaba Damo Academy launched a multi-modal large language model named Valley 2, designed for e-commerce scenarios. It aims to enhance performance across various fields and expand the application boundaries of e-commerce and short video scenarios through a scalable vision-language architecture. Valley 2 utilizes Qwen 2.5 as its LLM backbone, paired with the SigLIP-384 visual encoder, incorporating MLP layers and convolution for efficient feature transformation.

The latest release from the Alibaba team, mPLUG-Owl3 is a general-purpose multi-modal large model, with its core capability being the understanding of long image sequences. By introducing a hyper attention module, mPLUG-Owl3 can efficiently process visual and language information, achieving in-depth understanding and communication of multi-modal data such as images and videos. This model has made significant breakthroughs in inference efficiency, image processing capabilities, and the application of multi-modal knowledge, particularly in video understanding, where it can 'watch' a 2-hour movie in 4 seconds and accurately answer related questions.

The Sinan OpenCompass team at the Shanghai Artificial Intelligence Laboratory has collaborated with the Modao ModelScope to launch the Compass Multi-Modal Arena, a new section of a large model evaluation platform focusing on multi-modal large models. Users can upload images and input questions to have two anonymous multi-modal large models generate answers, which can then be subjectively evaluated based on the quality of the generated content, allowing users to select the better-performing model. The platform offers an easy-to-use interface and a unique question bank.

The "LingDong" brand has been meticulously crafted by NetEase Vuex, relying on its independently developed industrial-grade large models and AOP (Artificial Intelligence Operation Perception) technology concept. Two flagship products under the brand, the mining robot and the loading robot, have been put into use in over 50 provincial key projects, serving diverse environments such as mines, ports, mixing stations, and schools. In addition to the robot brand launch, NetEase Vuex also showcased it

Today, at the Huawei HDC2024 Developer Conference, Huawei Executive Director and Huawei Cloud CEO Zhang Pingan announced the official release of Huawei Cloud's Pangu Big Model 5.0 to the world. The Pangu Big Model 5.0, with its parameters ranging from hundreds of millions to trillions, along with its diverse functions including Pangu Natural Language Model, Multimodal Model, Visual Model, Predictive Model, and Scientific Computing Model, provides unprecedented robust support for AI applications.

Shengshu Technology has released a general multi-modal large model, along with tools such as PixWeaver and VoxCraft. The multi-modal large model adopts a foundational fusion approach to integrate multi-modal knowledge. PixWeaver is a visual creative design platform that supports the generation of images in various styles. VoxCraft is a tool that enables the rapid creation of 3D images. Shengshu Technology's large model also offers service capabilities across multiple industries, including game development, film production, and advertising marketing.

Kunlun Tech stated that the 'Tiangong' large model has been iterating on a weekly basis since its release, with the training cluster operating at high load. The mobile version of the Tiangong AI assistant has officially launched and entered the internal testing phase, available for both iOS and Android users to download and test. The company's large model supports text conversations exceeding ten thousand characters, allowing users to engage in over 20 rounds of interaction.