Alibaba Cloud Launches ‘ModelScope,’ An Open-Source Model-as-a-Service (MaaS) Platform that Comes with Hundreds of Artificial Intelligence (AI) Models

The Chinese technology conglomerate Alibaba Group’s subsidiary, Alibaba Cloud, serves as its technology and intelligence backbone. Founded in 2009, Alibaba Cloud aims to offer its customers all across the world a full range of cloud computing services, which include big data analytics, a machine learning platform, IoT services, network virtualization, storage, administration, and elastic computing. As a part of its annual Apsara Conference, the company announced the launch of  ModelScope, an open-source Model-as-a-Service (MaaS) platform that features hundreds of AI models, including sizable pre-trained models for international developers and researchers.

While changing the paradigm of software development and accelerating the integration of the cloud and endpoint terminals, cloud computing has caused a fundamental advancement in the way computer resources are produced, organized, and put to commercial use. To lower the work required for businesses to adopt new technology and maximize their prospects in the cloud era, Alibaba Cloud has worked on upgrading their cloud-based services and tools to become serverless and more intelligent as more customers begin using cloud-based services. To further assist clients in achieving business innovation through cloud technologies, the world’s top cloud provider enhanced its integrated data analytics and intelligent computing platform and unveiled various serverless database solutions during its flagship conference.

Over 300 ready-to-deploy AI models have just been added to the ModelScope platform. Alibaba’s global research program, the Alibaba DAMO Academy, created these models over five years. These models cover a range of domains, including audio processing, natural language understanding, and computer vision. The site also features more than 150 cutting-edge models that are renowned internationally as the best in their fields for a certain task. The platform also consists of large pre-trained models developed exclusively by Alibaba, such as Tongyi (which can convert text into images with five billion parameters) and One-For-All (a model with six billion parameters that excels at cross-modal tasks like captioning photos and answering visual questions). Independent developers have also added numerous models to the open-source platform.

ModelScope was developed to enable the development of AI models easily in a more affordable manner by allowing users to quickly and easily test models online for free. Programmers and researchers can create new AI applications by optimizing current models that can be operated on Alibaba Cloud or on other cloud computing services or locally. By making it easier for universities and smaller businesses to design and run AI models for their research and business, respectively, the launch highlights DAMO’s continued efforts and commitment to supporting transparent and inclusive technology. The community is anticipated to expand as more high-quality models soon become accessible on the platform from DAMO, partners from academic institutes, and independent developers.

Alibaba Cloud is making its core cloud products serverless thanks to recent technological developments in serverless software development, allowing customers to focus on product deployment and development without worrying about managing servers and infrastructure. The latest offerings from Alibaba Cloud concentrate on giving users access to processing capacity on demand. These goods include the cloud-native data warehouse AnalyticDB (ADB), the cloud-native database PolarDB, and ApsaraDB for Relational Database Service (RDS). Customers can benefit from automatic scaling with extreme flexibility based on actual workloads and a pay-as-you-go billing approach to cut costs using Alibaba Cloud’s serverless technology. Demands can have an automatic, elastic scaling time of as short as one second. Additionally, organizations can save their expenditures by an average of 50% by employing updated database goods instead of conventional ones.

Alibaba Cloud recently updated its ODPS (Open Data Platform and Services), a self-developed comprehensive data analytics and intelligent computing platform to provide businesses with various data processing and analytics services. The platform can manage online and offline data simultaneously in one system, giving firms with heavy workloads access to analytics for more effective and less expensive corporate decision-making. Alibaba Cloud also made the Wuying Architecture launch announcement with a presentation of its applications on the Wuying Cloudbook to promote workload collaboration between the cloud and local hardware. The Cloudbook with specialized Architecture is made to enable cooperation and flexibility at work by enabling users to access cloudless computing power more securely and quickly. With the introduction of ModelScope and other such cloud services that utilize serverless software development, Alibaba Cloud intends to improve product deployment and development.

Check out the reference article and tool. All Credit For This Research Goes To Researchers on This Project. Also, don’t forget to join our Reddit page and discord channel, where we share the latest AI research news, cool AI projects, and more.

[Announcing Gretel Navigator] Create, edit, and augment tabular data with the first compound AI system trusted by EY, Databricks, Google, and Microsoft