Lilt Announces Third-Party LLM Hub and Customized Self-Service Model Builder

Lilt宣布第三方LLM中心和定制自助服务模型生成器

2023-09-22 07:00 Lilt

本文共721个字,阅读需8分钟

阅读模式 切换至中文

New Contextual AI Platform enables custom model manipulation and advanced analytics for global enterprises • • • San Francisco, CA - September 14, 2023 - Lilt, the leading AI solution for enterprise translation, today at its third ever AI Day announced the company’s latest innovations, including third-party large language model integration in the Contextual AI Engine, self-serve model building and training, and advanced program analytics. As leading enterprises rush to implement a global AI and LLM strategy, Lilt’s Contextual AI Platform offers a quick and scalable path to achieving extraordinary ROI from AI. “As the speed of AI innovation accelerates, global enterprises are eager to look beyond the hype and build AI programs that they can implement in production with best-in-class technology,” said Lilt CEO Spence Green. “We’re thrilled that we can now offer our customers the ability to orchestrate and manipulate company-specific LLMs that they can quickly operationalize at scale.” During the 90 minute live virtual event, Lilt announced several new features across multiple layers of the Contextual AI Platform, including: LILT-V3 Models Following the release of LILT-V2 in Spring 2023, Lilt today released LILT-V3, its newest-generation LLMs that offer improved output accuracy. LILT-V3 leverages cutting-edge methods for fine-tuning and prompting, with in-context learning and a training data system that optimizes for high-relevance domain-specific data sets and context. The model’s larger neural networks synthesize and generalize this information and interpret those high-relevance examples in real-time for optimal model outputs. With these models in production, Lilt customers can now advance their AI strategy for global customer reach and achieve incredible quality at scale. Third-party LLM Support & Model Builder Lilt’s newest foundation model integration enables customers to harness Amazon Translate, DeepL, and Google Translate models right within the Lilt Platform. With this new offering, businesses can now seamlessly choose their large language models of choice, unlocking the best models for their unique enterprise use case. As leading enterprises develop bespoke LLM strategies, the Lilt Contextual AI Platform enables comprehensive execution of that strategy across workflows and business systems. This third-party LLM support pairs enterprise LLM preferences with Lilt’s proven production processes, linguists, and Translate tool - providing the highest quality Instant Translation for their use case. Model Builder Lilt additionally announced its Model Builder, a new management feature within the platform’s orchestration layer. Lilt Model Builder enables fine-tuning of third-party enterprise LLMs to be used in the Lilt Contextual AI Engine. The Lilt Platform has historically enabled model fine-tuning for Lilt’s proprietary LLMs, with a continuous training cycle that feeds training data to models in real time as human translators and verifiers perform work in the platform. With Model Builder, Lilt customers can now apply that same quality-verified training data to fine-tuning of third-party LLMs. Model Builder empowers Lilt customers to create purpose-built LLMs that generate customized, high quality, and brand aligned outputs for a range of business applications. Users can fine-tune their models in just a click with Lilt’s model building and training interface, optimizing model quality on the cadence of days or weeks instead of months or years. Advanced Analytics Lilt additionally announced new and improved premium program insights via a suite of advanced analytics tools that empower customers with in-depth insights into their Contextual AI program and operations, all within the Lilt Platform. This user-friendly tool offers customizable visualizations of vital metrics and program-level insights to support strategic decisions and optimization of the AI strategy, quality, and business impact. These new features are available to Lilt customers and are in production starting today, powering enterprise use cases across a range of industries and business units. For more information about Lilt and its AI-powered solutions, please reach out to contact@lilt.com. • • • About Lilt Lilt is the leading AI solution for enterprise translations. Our stack, made up of our Contextual AI Engine, Connector APIs, and human feedback, enables global organizations to adopt a true AI translation strategy, focusing on business outcomes instead of outputs. With Lilt, innovative, category-defining organizations like Intel, ASICS, WalkMe, and Canva are using AI technology to deliver multilingual, digital customer experiences at scale. Visit us at lilt.com or contact us at contact@lilt.com.
新的上下文人工智能平台为全球企业提供定制模型操作和高级分析 ••• 加利福尼亚州旧金山——2023年9月14日——领先的企业翻译人工智能解决方案Lilt今天在其第三个人工智能日宣布了该公司的最新创新,包括上下文人工智能引擎中的第三方大型语言模型集成、自助式模型构建和培训以及高级程序分析。随着领先企业急于实施全球人工智能和LLM战略,Lilt的上下文人工智能平台提供了一种快速和可扩展的途径,以实现人工智能的非凡投资回报。 Lilt首席执行官斯宾塞·格林表示:“随着人工智能创新速度的加快,全球企业都渴望超越宣传,建立人工智能程序,并利用一流的技术在生产中实施。”“我们很高兴我们现在可以为我们的客户提供编排和操作公司特定的有限责任管理系统的能力,他们可以快速大规模运营。” 在90分钟的现场虚拟活动中,Lilt宣布了跨上下文人工智能平台多层的几项新功能,包括: LILT-V3型号 继2023年春季发布LILT-V2之后,Lilt今天发布了LILT-V3,这是其最新一代的LLMs,提供了更高的输出精度。LILT-V3利用尖端的方法进行微调和提示,具有上下文学习和针对高相关性领域特定数据集和上下文进行优化的训练数据系统。模型的较大的神经网络综合和概括这些信息,并实时解释那些高相关性的例子,以获得最佳的模型输出。随着这些模型的生产,Lilt客户现在可以推进他们的人工智能战略,以覆盖全球客户,并实现令人难以置信的大规模质量。 第三方LLM支持和模型生成器 Lilt最新的基础模型集成使客户能够在Lilt平台内利用Amazon Translate、DeepL和Google Translate模型。借助这一新产品,企业现在可以无缝地选择他们选择的大型语言模型,为他们独特的企业用例释放最佳模型。随着领先企业开发定制的LLM战略,Lilt上下文人工智能平台能够跨工作流和业务系统全面执行该战略。 这种第三方LLM支持将企业LLM偏好与Lilt成熟的生产流程、语言学家和翻译工具相结合,为他们的用例提供最高质量的即时翻译。 模型生成器 Lilt还宣布了它的模型生成器,这是平台编排层中的一个新的管理特性。Lilt Model Builder支持在Lilt上下文AI引擎中使用第三方企业LLMs的微调。Lilt平台在历史上支持Lilt专有LLMs的模型微调,当人工翻译和验证人员在平台中执行工作时,连续的训练周期实时向模型提供训练数据。使用Model Builder,Lilt客户现在可以将相同的质量验证训练数据应用于第三方LLM的微调。Model Builder使Lilt客户能够创建专门构建的LLM,为一系列业务应用程序生成定制的、高质量的、与品牌一致的输出。使用Lilt的模型构建和训练界面,用户只需点击一下就可以微调他们的模型,以天或周的节奏优化模型质量,而不是几个月或几年。 高级分析 此外,Lilt还通过一套高级分析工具宣布了新的和改进的高级计划洞察,使客户能够深入了解他们的上下文人工智能计划和操作,所有这些都在Lilt平台内。这种用户友好的工具提供了重要指标和程序级见解的可定制可视化,以支持战略决策和人工智能战略、质量和业务影响的优化。 这些新功能可供Lilt客户使用,并从今天开始投入生产,为一系列行业和业务部门的企业用例提供支持。 有关Lilt及其人工智能解决方案的更多信息,请联系contact@lilt.com。 ••• 关于轻快 Lilt是领先的企业翻译人工智能解决方案。我们的堆栈由我们的上下文人工智能引擎、连接器API和人类反馈组成,使全球组织能够采用真正的人工智能翻译战略,专注于业务成果而不是产出。随着Lilt的出现,英特尔、ASICS、WalkMe和Canva等创新的、定义类别的组织正在使用人工智能技术来大规模提供多语言、数字客户体验。 访问我们在lilt.com或联系我们在contact@lilt.com。

以上中文文本为机器翻译,存在不同程度偏差和错误,请理解并参考英文原文阅读。

阅读原文