Deci, a deep learning company that harnesses AI to build AI, has unveiled the latest addition to its suite of innovative generative AI models, DeciLM-7B, a large language model with 7 billion parameters. Building on the success of its predecessor DeciLM 6B, DeciLM 7B sets new standards in the large language model (LLM) field, outperforming prominent open source models such as Llama2 7B and Mistral 7B in terms of accuracy and efficiency.
DeciLM-7B boasts unparalleled performance, outperforming open source language models by up to 13 billion parameters in terms of accuracy and speed with less computational demand. It achieves a 1.83x and 2.39x increase in throughput compared to the Mistral 7B and Llama 2 7B, respectively, which means much faster processing speeds compared to competing models. Its compact design is ideal for cost-effective GPUs, achieving an unparalleled balance between affordability and cutting-edge performance.
The impressive performance of the DeciLM-7B can be further accelerated when used in conjunction with Infery-LLM, the world’s fastest inference engine, designed to provide high throughput, low latency, and cost-effective inference on widely available GPUs. This powerful duo sets a new standard in throughput performance, achieving speeds 4.4x faster than Mistral 7B with vLLM without sacrificing quality. Leveraging DeciLM-7B in combination with Infery-LLM enables teams to significantly reduce LLM compute overhead, while simultaneously benefiting from faster inference times. This integration facilitates efficient scaling of generative AI workloads and supports the transition to more cost-effective hardware solutions.
This synergy allows efficient service to many clients simultaneously without excessive computational costs or response time issues. This is especially critical in sectors such as telecommunications, online retail, and cloud services, where the ability to respond to a massive influx of simultaneous customer inquiries in real-time can significantly enhance user experience and operational efficiency.
Licensed under Apache 2.0, DeciLM-7B is available for use and deployment anywhere, including on-premises, enabling teams to fine-tune specific industry applications without compromising data security or privacy. Its versatility allows teams to easily customize it for unique use cases across a wide range of business applications, including content creation, translation, conversation modeling, data classification and summarization, sentiment analysis, and chatbot development, among others. When fine-tuned for specific datasets, the DeciLM-7B can deliver quality similar to that of larger models such as GPT 3.5 at approximately 97% lower cost and better speed.
“With the increasing use of generative AI in various business sectors, there is a growing demand for models that are not only high-performance, but also cost-effective in operation,” said Yonatan Givman, CEO and co-founder of Deci. “Our latest innovation, DeciLM-7B, combined with Infery-LLM, is a game-changer in this regard. It is adaptable to diverse settings, including local solutions, and its exceptional inference efficiency makes high-quality large language models accessible.” A wider range of users.
The DeciLM-7B’s cost effectiveness and low computational demand make advanced AI technologies more accessible to businesses of all sizes, fostering innovation and driving digital transformation forward across sectors. With DeciLM-7B, companies can now harness the full potential of AI without the high costs or complexities previously associated with sophisticated language models.
Deci AI’s introduction of the DeciLM-7B builds on its proven track record of innovative and effective generative AI models, including the DeciLM 6B, DeciCoder 1B and DeciDiffusion 1.0. Similar to its other models, the DeciLM 7B is built using Deci’s advanced Automated Neural Architecture Architecture (AutoNAC) engine, the most advanced research-based neural architecture (NAS) technology on the market, with its focus on efficiency.
Sign up for the free InsideBIGDATA newsletter.
Join us on Twitter: https://twitter.com/InsideBigData1
Join us on LinkedIn: https://www.linkedin.com/company/insidebigdata/
Join us on Facebook: https://www.facebook.com/insideBIGDATANOW
Deci, a leading provider of deep learning AI infrastructure, has recently announced the release of DeciLM-7B, a revolutionary language model that offers a significant leap forward in performance and inference cost efficiency. This state-of-the-art model represents a major breakthrough in natural language processing, and has the potential to significantly impact various industries that rely on language modeling for their AI applications. With its innovative design and unparalleled capabilities, DeciLM-7B is poised to set a new standard for language model performance and cost-effectiveness.