experiment

Header

Footer

PreFooter

SectionTitle

PaginationButton

BlogPage

Trusted by world class companies, Scale delivers high quality training data for AI applications such as self-driving cars, mapping, AR/VR, robotics, and more.

AI & ML Blog | Scale AI

NextImage

AnnouncementContextProvider

Screen

scale.com (Production)

lottie-react

plasmic-nav

react-scroll-parallax-global

plasmic-embed-css

plasmic-query

plasmic-basic-components

plasmic-tabs

plasmic-cms

react-chartjs-2

plasmic-strapi

react-slick

radix-ui

react-quill

Customers

Product

Company

Engineering

Today we are excited to share several key updates that build on Scale AI’s mission to accelerate the development of AI. First, we are officially announcing Scale’s Series D funding round. Following our Series C fundraise last year, Scale has raised $155M at a post-money valuation north of $3.5B led by Tiger Global. We are in the midst of one of the most impactful technology transformations of human history with AI. The potential applications of AI are going to deeply improve the world: scalable medical diagnosis to everyone in the world, massive reduction in driving accidents, goods which are dramatically cheaper to manufacture and transport, and much more. What’s holding us back from building those dreams is lack of infrastructure. While many organizations are adopting AI, they are forced to build their technology stack from scratch, significantly slowing progress. Four years ago, we recognized this problem and sought to unlock the bottlenecks holding back AI development. We started by building a platform that solves the necessary, foundational layer that fuels all machine learning models: large annotated datasets. Our solution gives developers the ability to send data to our platform through the Scale API and receive high-quality training data that’s ready for building production-grade AI algorithms. We use a combination of machine learning and human intelligence to annotate tens of millions of data points per week for hundreds of customers like OpenAI, DoorDash, Pinterest, General Motors, NVIDIA, Nuro, and Toyota. We’ve chosen to stay focused on the applications of AI with the greatest potential for impact: autonomous driving, government, and cutting-edge consumer technology. As AI has transformed every industry, we’ve continued to see demand for our platform to grow—from e-commerce and AR /VR to relevance &amp; personalization and financial services. We are still very early in the journey of AI, and we remain optimistic about all the future dreams we will support. Our team is tackling the next bottleneck in AI development: managing the full life cycle of AI development across teams. Building and deploying scalable machine learning infrastructure is slow, manual and costly, and in August we launched <a href="https://scale.com/nucleus" rel="noopener noreferrer" target="_blank">Nucleus</a> to enable seamless collaboration across the AI development cycle. Teams can visualize and organize their datasets, test and compare models, and share data with colleagues. With Nucleus, our goal is to become the mission control for your data, and over time we’ll be rolling out support for more data types and adding more features to make developing AI easier and faster than ever.<img src="https://imagedelivery.net/wLbZE4_NzVVdgHc15St55g/0620751d-5f96-4b5f-0469-e11b65805a00/public" alt="Scale_Nucleus_ML_Development_Cycle"> As part of today’s announcement, we are also excited to announce the acquisition of Helia AI. Helia’s team brings deep expertise in applying AI to real-time video, from core algorithm development to necessary AI infrastructure. Their AI infrastructure experience aligns closely with what we’re building with Scale Nucleus, and the Helia team will help advance the Scale platform as we expand its capabilities. We look forward to continuing our work of building AI infrastructure with Scale’s hundreds of customers, and our new funding will allow us to accelerate hiring, expand to new markets and build more robust products to match the evolving needs of AI as the technology matures. We’re still in the early days of AI, and we believe Scale will enable a future where AI is as easy to build and manage as any other software. If you are interested in joining us on our mission, check out our <a href="https://scale.com/careers" rel="noopener noreferrer" target="_blank">open positions</a>. If you’re interested in partnering with us, please <a href="https://scale.com/sales" rel="noopener noreferrer" target="_blank">reach out</a>.

Accelerating the development of AI applications and empowering AI dreamers with a new round of funding led by Tiger Global. 

Scale AI's Series D: Expanding our team to empower AI dreamers

Scale AI’s Series D: Expanding our Team to Empower AI Dreamers

When Covid-19 first hit, many companies had to rapidly reorient their operations – from altering their supply chains to implementing social distancing and work-from-home orders. One sector that was particularly hard hit was the restaurant industry. When in-person dining was suspended, restaurants turned to delivery to survive. But setting up a delivery service from scratch takes time and money that many establishments, with their income drying up, did not have. Suddenly, food delivery platforms became an economic lifeline. To ensure restaurants could onboard quickly and seamlessly in these critical early days, one of the leading delivery platforms turned to Scale.<h3>Scaling labeling</h3> Every time a restaurant joins one of the delivery platforms or changes its menu, the menu’s contents need to be inputted into the platform so users can select their food. Inputting menus manually is slow, expensive, and leads to mistakes. With restaurants relying on delivery to reach customers and support their staff, a leading delivery platform knew it was vital to ensure an efficient process for its partner restaurants. This presented Scale with unique latency and quality control challenges, requiring both an incredibly high level of accuracy with a fast turnaround to provide operational efficiencies at a critical time for our customer. Our aim was to use the combination of partial automation and human quality control in our <a href="https://scale.com/document" rel="noopener noreferrer" target="_blank">Scale Document</a> labeling pipeline to make dramatic efficiency improvements over manual processes. In all, we nearly halved the time it takes our customer to process menus and reduced critical error rates to &lt;1% for all items labeled, allowing them to onboard restaurants smoothly exactly when they needed it.<h3> </h3><h3> Building the right data infrastructure</h3> While processing standardized documents such as food menus, government IDs or loan application forms might be intuitive and simple for humans, it is a surprisingly nontrivial problem for algorithms, involving much more than automatic transcription. In particular, the data often contains many dependencies that need to be captured accurately. For example, ordering a pepperoni pizza doesn’t just mean choosing the type, but also the size, the type of base, choosing to substitute or customize toppings (which can also vary in price depending on the size of the pizza, creating more interdependencies), and whether to add extras like burgers or sandwiches. It’s important to capture these dependencies in the menu correctly, otherwise inputting errors risk causing losses for both the restaurant and the delivery platform.<img src="https://imagedelivery.net/wLbZE4_NzVVdgHc15St55g/e10f107d-1efe-45c5-45a3-db2291017800/public" alt="DD_decision"> Algorithmically, this has to be represented as complex decision trees made up of categories, items, and lists of options, capturing the relationships between them. Thankfully, both our document processing pipeline and form capture tools were already set up to capture these types of nested data structures automatically without the need for customization. This meant we were able to quickly pivot our tools for our customer’s requirements. <h3>Judicious Automation and Scaling</h3> Once we have the right tooling in place to pull the raw data and label complex data structures, we can then start processing the data. Given how many menus delivery platforms were processing at the start of shelter-in-place, and how important for restaurants’ continued operations it was to process this data quickly with no drop-off in quality, we need to make this process as efficient as possible. At Scale AI, we believe that a mix of judicious labeling automation with continued human oversight is the only way to provide data at the scale, quality, and low cost needed to enable many enterprise applications of AI. In particular, we aim to automate the lower-skilled parts of the data labeling pipeline to focus human review on the hard parts – quality control, edge cases, and the most complex data types. That way, we can help guarantee quality without sacrificing efficiency. We use automation to prefill menu taxonomies for expert human labelers to review and confirm. We trained our form capture tool to automatically understand document structure and predict the next transcription. And when our labelers needed to add inputs manually, we built a “smart suggestion” feature that provides them with ready prompts for common items that they have already encountered in their batch of menus. Both have helped improve both the efficiency and accuracy of labeling simultaneously: allowing us to transcribe more than 3,000 menus in a single day and return menus within short turnaround times to onboard restaurants within 24 hours from start-to-finish. By drawing on our wide network of well-trained labelers and augmenting their work with careful automation, we’re able to handle changes in demand dynamically – essential for providing a smooth service during sudden peaks in demand. <h3> </h3><h3> Custom Benchmarks and Integrations</h3><h3> </h3> Once data is flowing smoothly and efficiently through this labeling pipeline, the next step is to ensure that the data is being labeled to an incredibly high standard, with almost zero errors. We’ve developed a range of techniques to guarantee label quality, including the randomized use of benchmark quality tests, confidence-based consensus, and, for ambiguous labeling requirements, our own <a href="https://scale.com/blog/how-to-label-1m-data-points-week" rel="noopener noreferrer" target="_blank">automated benchmark generation system</a>. Benchmarking is not always an objective science – the most effective benchmarks are often task-specific. The high stakes of onboarding restaurants during a pandemic required particularly high guarantees of labeling quality. To meet them, we worked closely with our customer to build custom benchmarks for this workflow, turning the most complex mislabeled menus into benchmarks that assessed our labelers’ performance. In surveying mislabeled menus, we noticed two key groups of errors. One was associated with the cross-referencing of menu sections between many items and mistaking how they relate to each other, while the other group tended to be mistakes from larger menus where it may be less obvious when certain features of the menu are missing. By identifying these key groups and creating benchmarks out of the associated menus, we can ensure that tasker performance in such cases exceeds the necessary standard to meet menu quality needs. <img src="https://imagedelivery.net/wLbZE4_NzVVdgHc15St55g/fc0a9a32-08be-4505-422b-a05d016c3f00/public" alt="Menu_dd_photos"> The menu on the left showcases large amounts of cross-referencing and item optionality, while the menu on the right shows a fairly common amount of menu feature density. <h3>What’s Next</h3> AI models are rapidly becoming critical infrastructure for a huge range of businesses – a trend of automation that the COVID pandemic has rapidly accelerated. But gathering effective, accurate, and unbiased data currently remains such a challenging task that it is preventing real-world AI systems from reaching their full potential and presenting high complexity barriers for smaller companies. We’re tackling a whole host of problems to help accelerate that progress. Right now, teams at Scale AI are making better use of machine learning to make labeling orders of magnitude more efficient, building tools to process increasingly complex data types, such as 3D point clouds in computer vision, and developing new infrastructure tools to help streamline the management of data. Our first such management tool, <a href="https://scale.com/blog/introducing-scale-nucleus" rel="noopener noreferrer" target="_blank">Scale Nucleus</a>, is already helping our computer vision customers automate time-consuming manual steps in the ML development process. We’re now working on deploying it for our customers in natural language. Ultimately, we want to make it as easy to deploy AI as any other type of software. If you’re interested in joining us in solving these problems, take a look at our <a href="https://scale.com/careers" rel="noopener noreferrer" target="_blank">careers page</a> for our latest open positions. If you have projects that require high-quality data labeling, <a href="https://scale.com/sales" rel="noopener noreferrer" target="_blank">let our team know</a>!

How Scale AI worked with a leading food delivery platform to ensure restaurants could onboard quickly and seamlessly. 

Rapidly scaling menu transcription to support restaurants during the pandemic

Scaling Menu Transcription Tasks with Scale Document

The Scale AI team hosted its first ever Hackathon last month. The hackathon was designed to:<ol><li> </li><li> Encourage innovative thinking to solve some of Scale’s biggest challenges</li><li> </li><li> </li><li> Prototype and potentially launch impactful projects that make Scale better</li><li> in practice</li><li> </li><li>Facilitate long-term cross-functional relationships</li><li>Give all Scaliens a creative break from their daily work.</li><li> </li></ol> Scaliens across functions and teams leapt to the challenge, proposing 63 ideas and ultimately working on 10 separate projects comprising of 37 team members. Projects ranged from new product proposals to better internal search, from refactoring tech debt to streamlining instructions for Taskers.<img src="https://imagedelivery.net/wLbZE4_NzVVdgHc15St55g/e7572ed8-8b09-48ee-088a-063ab2078f00/public" alt="Nofel_Naoman_Hackathon_Quote"> <h2>Project Highlights:</h2> The following are just some of the projects Scaliens tackled during the Hackathon. As a disclaimer, most of these projects are not live, nor do they indicate immediate plans to build these ideas. <h3>Super Instructions</h3>Winner: Best In ShowTeam Members:<a href="https://www.linkedin.com/in/svanweelden/" rel="noopener noreferrer" target="_blank">Shaun VanWeelden</a>, <a href="https://www.linkedin.com/in/taisiakaraseva/" rel="noopener noreferrer" target="_blank">Taisia Karaseva</a>, <a href="https://www.linkedin.com/in/jeffrey-linj/" rel="noopener noreferrer" target="_blank">Jeff Lin</a>, <a href="https://www.linkedin.com/in/galenbertozzi/" rel="noopener noreferrer" target="_blank">Galen Bertozzi</a>, <a href="https://www.linkedin.com/in/rashidlasker/" rel="noopener noreferrer" target="_blank">Rashid Lasker</a>, <a href="https://www.linkedin.com/in/ivan-roumec-b0450a6b/" rel="noopener noreferrer" target="_blank">Ivan Roumec</a>, <a href="https://www.linkedin.com/in/nofel-naoman-aa28a0142/" rel="noopener noreferrer" target="_blank">Nofel Naoman</a>, <a href="https://www.linkedin.com/in/nba/" rel="noopener noreferrer" target="_blank">Lawder Paul</a>Problem Statement: Annotation instructions are critical for the development of high-quality training data. Today’s annotation instructions are mostly made on static, non-interactive documents, not connected to the API payloads creating the task and requiring 3rd party solutions to host. Solution: The Super Instructions team tackled this problem in 3 ways: <ol><li> </li><li> Extending the Task Schema used in our APIs to accommodate labeling</li><li> instructions.</li><li> </li><li> </li><li> Creating an in-app instruction builder that simultaneously creates the</li><li> instructions as well as the API payloads that would need to be leveraged</li><li> in task creation.</li><li> </li><li> </li><li> Bringing a world-class interactive design to the in-app educational</li><li> experience for those doing the annotating, including features like search,</li><li> navigation, and zoomable images.</li><li> </li></ol> The goal of this prototype was to re-imagine the experience both our customers and Taskers have when creating and completing work on Scale’s platform. We firmly believe better instruction design and more consistency leads to higher-quality annotations for all involved. <h3>Corp Search</h3>Winner: Audience ChoiceTeam Members:<a href="https://www.linkedin.com/in/nba/" rel="noopener noreferrer" target="_blank">Lawder Paul</a> and <a href="https://www.linkedin.com/in/james-lennon-30421078/" rel="noopener noreferrer" target="_blank">James Lennon</a>Problem Statement: “Corp” is Scale AI’s admin portal for employees to gather information on on-going projects and to influence the quality and outcome of data annotation. The Scale AI team has launched countless tools and introduced new functionalities to Corp to improve our operations but the Corp portal itself was not very user friendly. There are now so many tools in Corp that it can be hard to find the exact tool, and for new Scaliens it’s not always clear what different tools do. <img src="https://imagedelivery.net/wLbZE4_NzVVdgHc15St55g/6a652d50-f749-461b-3c7a-dc221ec55600/public" alt="Scale Corp Search"> Solution: Corp Search allows employees to quickly search across all areas and tools within Corp, packaged in an elegant "command-line" interface that can be surfaced anywhere in Corp. <h3>Corp Migration</h3>Winner: Most Immediately ImpactfulTeam Members:<a href="https://www.linkedin.com/in/diegomura/" rel="noopener noreferrer" target="_blank">Diego Muracciole</a>, <a href="https://www.linkedin.com/in/federicomarcos/" rel="noopener noreferrer" target="_blank">Federico Marcos</a>, <a href="https://www.linkedin.com/in/martin-cabral-b0897615/" rel="noopener noreferrer" target="_blank">Martin Cabral</a>Problem Statement: As stated in the Corp Search project, Corp had become a complex and sophisticated tool used by many to provide the best service to our customers. Many of the Tasker and customer-facing features and tools were tied together or interlinked with Corp, making additional engineering work more challenging. Solution: The team spent the week separating these Tasker and customer-facing features were separated from Corp. The team also developed a framework to encourage internal modularity and code reuse to enable faster deployment in the future as Scale continues to grow. The team also ensured backwards compatibility with previous code to ensure daily operations could continue. This project has already enabled several refactors and performance improvements. <h3>The Great Refactor</h3>Team Members:<a href="https://www.linkedin.com/in/liuandrewk/" rel="noopener noreferrer" target="_blank">Andrew Liu</a>, <a href="https://www.linkedin.com/in/nick-induni-44357292/" rel="noopener noreferrer" target="_blank">Nick Induni</a>, <a href="https://www.linkedin.com/in/steven-hao-160b9638/" rel="noopener noreferrer" target="_blank">Steven Hao</a>Problem Statement: Like many fast-growing startups, Scale AI has accumulated some level of tech debt in order to quickly deliver new features and products to our customers. Solution: The Great Refactor team looked to tackle this issue by investing a week of their time to improve code quality to optimize for long-term maintainability and development velocity. <h3>Tasker Scoresheets</h3>Team Members:<a href="https://www.linkedin.com/in/akshat-bubna-188885103/" rel="noopener noreferrer" target="_blank">Akshat Bubna</a>, <a href="https://www.linkedin.com/in/rilka/" rel="noopener noreferrer" target="_blank">Rilka Li</a>, <a href="https://www.linkedin.com/in/alan-yu-954926110/" rel="noopener noreferrer" target="_blank">Alan Yu</a>Problem Statement: Although Taskers (our global data annotation workforce) regularly receive feedback on the quality of their submissions, it is difficult to provide granular feedback with specific areas of strength and weakness. Solution: The team developed Scoresheets that could be accessed on the Tasker dashboard. The dashboard allows Taskers to track their individual performance on specific skills over time and provides learning resources to improve at weak skills. <h3>Scale Grep</h3>Team Members:<a href="https://www.linkedin.com/in/jonfung1/" rel="noopener noreferrer" target="_blank">Jonathan Fung</a>Problem Statement: Instead of searching through command-line grep, IDE search, or Github search, which can be slow and unwieldy, engineers need quick, interactive, on-the-fly code search that can generate quickly sharable links. Solution: We deployed two open-source projects, Livegrep and Webgrep internally, and added automated index-generation so that the code in our git repositories gets indexed every 2 hours. Now, engineers can search through our large codebase in milliseconds and click code fragments to generate sharable links to exact lines of code. <h2> </h2><h2> Facilitating Interactions to Foster Cross-Functional Relationships</h2><h2> </h2> In addition to working on impactful projects, another goal of the Hackathon was to facilitate cross-functional relationships. Fostering these connections when everyone is remote can be particularly tricky. How do you recruit teammates to work on your project when you can’t organically run into folks in hallways? How do you whiteboard ideas, or learn a new skill from teammates when you can’t physically sit together? <h3>Pitch Session</h3> To help idea originators find teammates, we held a company-wide pitch session a week before the Hackathon. Scaliens then spent the following week signing up for the pitched projects and forming cross-functional teams. <h3>Events and Activities</h3> <img src="https://imagedelivery.net/wLbZE4_NzVVdgHc15St55g/9e1ad977-b0ea-4bc4-1903-e9cc346ab700/public" alt="Scale_Hackathon_Zoom"> To help Scaliens develop new skills, we organized workshops throughout the week of the Hackathon. Workshops included: <ul><li>A session on Storybook for planning engineering deployments</li><li> </li><li> A session on Figma to understand Scale’s design resources and to enable</li><li> better designing and prototyping</li><li> </li><li>A session on Tableau to enhance data visualization.</li></ul> We also organized daily social activities and special events throughout the week to facilitate social bonding. These activities included a Daily “Brain Game” and a virtual Game Night. To help everyone get into the spirit of the Hackathon (whether they were on a team or not), we fully branded the hackathon experience. Our Brand Director, <a href="https://www.linkedin.com/in/tywilkins/" rel="noopener noreferrer" target="_blank">Ty Wilkins</a>, created custom Zoom backgrounds, and branded swag, while Product Designer, <a href="https://www.linkedin.com/in/chun-jiang-0428/" rel="noopener noreferrer" target="_blank">Chun Jiang</a>, designed our Paranormal Hacktivity mascot. <img src="https://cdn.builder.io/api/v1/image/assets%2Fe0438815ba51486bbb6a202747122d4b%2F2c8bc6da94334d25a8882cfda83bdcfe" alt="scale_blog_hackathon_ghost"> <h3> </h3><h3> Presentations and Hackathon Wrap-up</h3><h3> </h3> Once projects were wrapped up, the teams presented their projects to the entire company. A team of internal judges gave out awards for “Best-In-Show” and “Most Immediately Impactful” and the entire company voted on a “Audience Choice” award. After the final presentations, Scaliens gathered in a virtual stretch session to unwind and relax. We hope this blog inspires you to host your own virtual hackathon or spark ideas for a hackathon that is in the works. If you’re interested in tackling new challenges and joining a collaborative and exciting team, take a look at all of our <a href="https://scale.com/careers#jobs" rel="noopener noreferrer" target="_blank">open positions</a>!

Encouraging innovative thinking and making Scale AI better in practice with our first virtual Hackathon. 

Encouraging innovative thinking and tackling Scale AI's biggest challenges.

Scale Hackathon

Back on September 16th, we introduced you to our <a href="https://scale.com/blog/introducing-our-new-api-docs" rel="noopener noreferrer" target="_blank">new API Docs</a>. Today, we’re pleased to follow up the original announcement with further details on how we implemented this with Next.JS, Tailwind CSS, and ReadMe. Furthermore, rather than simply describing how we used each platform and framework, we have packaged together an example for you to implement and build off of <a href="https://github.com/scaleapi/custom-documentation-example" rel="noopener noreferrer" target="_blank">in this repository</a>.<h2>Some Context:</h2> At Scale AI, we are building the critical tools that make it easier to build and scale real-world AI systems. To provide our customers with the best experience possible on the Scale AI platform, we believe that a clear, simple, and memorable front-end, user interface, and user experience is critical. Scale’s design team has been hard at work building out a design system and improving customer-facing interfaces to bring front-end consistency across our platform. Most important to this effort was to implement the design system across our different applications. We decided to use Tailwind CSS as a CSS framework and CSS Modules as a way to add custom elements when some customization was needed. To ensure our website and applications were highly performant and loaded quickly, we chose to use Next.js as our react framework.<h2>Why Tailwind?</h2> Scale’s design team have used Next.js and styled-jsx as a way of providing CSS styling since the initial launch of our website. While this was fine when there were fewer pages on the website, it proved to be hard to scale and maintain a design system without proper documentation. At this point the team faced two options: we could either spend a lot of time creating and improving our own CSS and writing documentation for it, or we could use a CSS framework that could adapt to our needs. Since the Scale team prefers not to re-invent the wheel when it isn’t needed, we decided that using a CSS framework would yield the best results in the shortest period of time. Tailwind has allowed us to expedite the development of new layouts and pages while maintaining consistency across our website. Tailwind also has the advantage of having strong documentation behind it to support future development. Using Tailwind has also freed the team up to spend more time focusing on special details for each page like custom animations and interactions.<h2>Why Next.js?</h2> Next.js is one of the best React Frameworks out today. Since the launch of <a href="https://nextjs.org/blog/next-9-3#next-gen-static-site-generation-ssg-support" rel="noopener noreferrer" target="_blank">SGG support</a> in March, it’s a no brainer for a fast and performant website and API docs. Quoting them from their own <a href="https://nextjs.org/blog/next-9-3#next-gen-static-site-generation-ssg-support" rel="noopener noreferrer" target="_blank">blog post</a>:<img src="https://imagedelivery.net/wLbZE4_NzVVdgHc15St55g/4a743efa-0436-43a8-18b6-9368be8e8a00/public" alt="Next.js Quote"> Next.js’s API routes also proved to be helpful as we used this to retrieve the search values made by the user. More on their API Routes <a href="https://nextjs.org/docs/api-routes/introduction" rel="noopener noreferrer" target="_blank">here</a>. <h2>Why Readme?</h2> As we highlighted in our previous <a href="https://scale.com/blog/introducing-our-new-api-docs" rel="noopener noreferrer" target="_blank">blog post</a>, we chose ReadMe as the platform for our API documentation because ReadMe allowed us to: <ol><li> </li><li> Unify the API reference and the Customer Dashboard documentation into a</li><li> unified website.</li><li> </li><li> </li><li> Improve the developer experience. Instead of only showing static code</li><li> examples, ReadMe enables developers to submit API requests directly with</li><li> their Scale API key. ReadMe also has built in search functionality to make</li><li> it easier for developers to find specific topics.</li><li> </li><li> </li><li> Democratize the content creation process. Instead of editing raw markdown,</li><li> making a pull request, and then deploying to a server, we just had to</li><li> update the text we needed in the ReadMe UI. Ultimately, the ReadMe</li><li> platform provided a richer and more dynamic user experience.</li><li> </li></ol><h2> </h2><h2> Sharing our example code with the developer community</h2><h2> </h2> We developed the new API documentation during Scale’s Hackathon. We are incredibly proud of what we accomplished in just one week and we wanted to share our solution with anyone who is looking to solve a similar problem. Rather than simply providing a step-by-step instruction for this solution, we have packaged together an example for you in our <a href="https://github.com/scaleapi/custom-documentation-example" rel="noopener noreferrer" target="_blank">Github repo</a>. We hope this sample code allows you to build your own beautiful and customized documentation.

How to customize your API Documentation’s front-end leveraging cutting-edge platforms and frameworks.

Customizing API Documentation with Next.JS, Tailwind CSS, and ReadMe.

Quick Start Guide to Customized API Documentation

<h2>Introduction</h2> The ML Team at Scale hosts a weekly reading groups where members choose papers from the broad AI/ML community and discuss them ranging from topics in Computer Vision to NLP to Active Learning. Here, we describe a brief summary of some of the insights we gained from various papers and how we aim to use some of that knowledge in future research projects and applications to Scale AI’s business.<h3><a href="https://arxiv.org/abs/2008.09049" rel="noopener noreferrer" target="_blank">Discovering Useful Sentence Representations from Large Pretrained Language</a></h3><h3><a href="https://arxiv.org/abs/2008.09049" rel="noopener noreferrer" target="_blank"> Models</a></h3>Presenter:<a href="https://www.linkedin.com/in/nishantsubramani/" rel="noopener noreferrer" target="_blank">Nishant Subramani</a> This is Scale’s first research paper and we wrote a <a href="https://scale.com/blog/discovering-useful-sentence-representations" rel="noopener noreferrer" target="_blank">blog post summarizing the paper</a>. This paper focuses on whether we can adapt pre-trained language models off-the-shelf as universal decoders. To be considered “universal,” a decoder must have an implicit representation for any target sentence such that it can recover that sentence exactly when conditioned on its representation. We investigated whether such representations exist and whether they can be easily discovered. Experiments show that not only do these representations exist for sentences from a variety of different genres, but also that our methods can recover these sentences almost perfectly without fine-tuning the underlying language model at all.<img src="https://imagedelivery.net/wLbZE4_NzVVdgHc15St55g/733c8e61-0b1b-4a08-9092-43c494a9fc00/public" alt="Scale blog ml papers B_01"> <h3><a href="https://arxiv.org/abs/1910.04126" rel="noopener noreferrer" target="_blank">Scalable Nearest Neighbor Search for Optimal Transport</a></h3>Presenter:<a href="https://www.linkedin.com/in/anastasiia-alokhina-534b1014a/" rel="noopener noreferrer" target="_blank">Anastasiia Alokhina</a> The authors of this paper look at the problem of nearest neighbor search with respect to Wasserstein distance. As Wasserstein distance has become popular in high-resource data environments and domains (images, text, etc.), search becomes prohibitively slow. As a result, approximation methods are important. The authors introduce a Quadtree variant called a Flowtree that formally shows better asymptotic accuracy and empirically show that it beats existing methods on a variety of different real-world datasets on either running time or accuracy. Flowtree is a fast nearest neighbor search algorithm for Optimal Transport (Wasserstein-1 distance), which has a linear running time and speeds up SOTA performance by up to 7x. They do this by evaluating the optimal tree flow in the original metric space. <img src="https://imagedelivery.net/wLbZE4_NzVVdgHc15St55g/2f1c10db-58ad-4966-b9aa-a69c64080500/public" alt="scale blog ml papers B_02"> At Scale AI, we work with lots of high dimensional data such as documents and images for a variety of tasks. For these, we often need to find good representation spaces for these data points such that we can do fast and highly accurate similarity search in order to find outliers in datasets in order to find errors in labeling and find examples to label to improve our models using active learning. We think the flowtree method here could be invaluable in giving us the ability to use a different distance for this end and is something we will explore going forward. <h3><a href="https://demuc.de/papers/schoenberger2016sfm.pdf" rel="noopener noreferrer" target="_blank">Structure-from-Motion Revisited</a></h3>Presenter:<a href="https://www.linkedin.com/in/ppyht2/" rel="noopener noreferrer" target="_blank">Hongbo Tian</a> Structure-from-Motion(SfM) is a field of computer vision techniques which aim to reconstruct full 3-dimensional scenes from only a series of images. Accurate SfM algorithm can reproduce colored point clouds similar to LiDAR. The authors of this paper along with <a href="https://colmap.github.io/bibliography.html" rel="noopener noreferrer" target="_blank">Schönberger’s thesis</a> introduced a robust computational pipeline for accurate dense 3D reconstruction. The authors also introduced <a href="https://colmap.github.io/index.html" rel="noopener noreferrer" target="_blank">COLMAP</a> - an open-sourced CUDA accelerated software to reproduce the results. <img src="https://imagedelivery.net/wLbZE4_NzVVdgHc15St55g/62e7cdd5-e0b4-4fc0-7dee-619984c25800/public" alt="Scale blog ml papers B_03"> At Scale AI, our labelling pipeline is not subjected to the same online computational constraints compared to our customers. This allows us to pursue batch methods to enrich our 2D dataset, enabling our human labelers to capture 3D contexts from only images. <h3><a href="https://storage.googleapis.com/pub-tools-public-publication-data/pdf/59f3bb33216eae711b36f3d8b3ee3cc67058803f.pdf" rel="noopener noreferrer" target="_blank">Representation Learning for Information Extraction from Form-like</a></h3><h3><a href="https://storage.googleapis.com/pub-tools-public-publication-data/pdf/59f3bb33216eae711b36f3d8b3ee3cc67058803f.pdf" rel="noopener noreferrer" target="_blank"> Documents</a></h3><h3><a href="https://storage.googleapis.com/pub-tools-public-publication-data/pdf/59f3bb33216eae711b36f3d8b3ee3cc67058803f.pdf" rel="noopener noreferrer" target="_blank"> </a></h3>Presenter: <a href="https://www.linkedin.com/in/nishantsubramani/" rel="noopener noreferrer" target="_blank">Nishant Subramani</a> The authors aim to extract structured information from form-like documents. They observe that forms have fields that often correspond to well-understood types and are often associated with a specific key phrase that bears a visual relationship with it. They also observe that key phrases in documents are drawn from a small vocabulary. Due to these observations, the authors build a system with two parts: candidate generation and scoring &amp; assignment. In the candidate generation phase each field type is associated with a candidate generator based on a cloud-based entity extraction service. Spans are then detected. In an invoice there are multiple dates. Every date in an invoice becomes a candidate for every date field in the target schema. This process is repeated for every target field. In the scoring and assignment phase, the goal is to find the correct extraction candidate for each field. To do this, a score is computed for each candidate independently using a neural model. Next, the most likely scored candidate is assigned for each field. This process yields an independent learned representation based only on the candidates’ neighborhood (this is entirely independent of other candidates of fields). The neural scoring model is below. <img src="https://imagedelivery.net/wLbZE4_NzVVdgHc15St55g/18c49c25-c170-4dcd-e6da-51144e415600/public" alt="Scale blog ml papers B_04"> Our major takeaways are that structured extraction is not super well-studied by the academic community, but is relatively simple. The extraction system presented has promising accuracy and generalizes well to different domains. The representations have some interpretability as well. Ultimately some of the learnings from this paper could influence some of our work on <a href="https://scale.com/document" rel="noopener noreferrer" target="_blank">Scale Document</a>. <h3><a href="https://arxiv.org/abs/1912.11370" rel="noopener noreferrer" target="_blank">Big Transfer (BiT): General Visual Representation Learning</a></h3>Presenter: <a href="https://www.linkedin.com/in/felix-lau-0386641a/" rel="noopener noreferrer" target="_blank">Felix Lau</a> This paper focuses on how to transfer pre-trained representations to other tasks in the visual space. They propose a recipe called Big Transfer (BiT) and achieve very strong performance on ILSVRC-2012 (87.5% top-1 accuracy), CIFAR-10 (99.4%), and the Visual Task Adaptation Benchmark (VTAB; 76.3%). <img src="https://imagedelivery.net/wLbZE4_NzVVdgHc15St55g/e2a0b8c1-282d-4511-835f-a183b368ad00/public" alt="Scale blog ml papers B_05"> BiT has two phases: upstream and downstream. In upstream pre-training, the authors first investigate the scale of computational budget and how that affects performance. Secondly, they look at ground normalization and weight standardization, which helps significantly with small-batch and large-batch training alike. In downstream transfer, the authors propose a cheap fine-tuning methodology by using the BiT-HyperRule to select the most important hyperparameters for tuning as a function of the task’s intrinsic image resolution and number of datapoints. The hyperparameters they thought were important were training schedule length, resolution, and whether to use MixUp regularization. There are a few takeaways from this method we found interesting. There’s a lot of value in balancing the computational budget and simplifying the experimental pipeline when needed. The BiT-HyperRule focusing on only a few hyperparameters was illuminating. We were interested in the dynamics of how large batches, group normalization, and weight standardization interplayed and were surprised at how poorly batch normalization performed relative to group normalization and weight standardization for large batches. The empirical strength of group normalization and weight standardization surprised us as well and we will start exploring these techniques in some of our computer vision models with lots of data available. However, the most impressive part was the empirical few-shot learning performance relative to the baseline. We are trying to build models that quickly adapt to new customers and domains and often struggle with fine-tuning, so these ideas are very top of mind. <h3><a href="https://arxiv.org/abs/1905.03677" rel="noopener noreferrer" target="_blank">Learning Loss for Active Learning</a></h3>Presenter: <a href="https://www.linkedin.com/in/rishgoyell/" rel="noopener noreferrer" target="_blank">Rishab Goyal</a> The authors present an active learning method that is a task-agnostic by attaching a loss prediction module to a target network and train it to be able to predict target losses of unlabeled inputs. The module can then provide insight into data that the model could predict incorrectly and thus improve the overall model by choosing just those examples to go and label. <img src="https://imagedelivery.net/wLbZE4_NzVVdgHc15St55g/105e736f-2bce-4ad5-6e99-7279ab964900/public" alt="Scale blog ml papers B_06"> Their general process is that for every labeled example you predict the target and predicted loss for your prediction. Next, you take the unlabeled pool and pass them through the model to get predicted losses. Lastly you annotate the top-k data points and add them to the labeled training set. The loss prediction module uses a ranking loss. As a data labeling company, we are very interested in active learning and how we can leverage our own strengths to better build ML models. We really liked the approach because it is widely applicable, but were concerned that there could be many degenerate cases. We weren’t convinced that the loss prediction module would be strong enough to offer significant improvements over uncertainty or coverage based methods such as entropy, query-by-committee, and coreset based methods. <h3><a href="https://arxiv.org/pdf/1912.13318v5.pdf" rel="noopener noreferrer" target="_blank">LayoutLM: Pre-training of Text and Layout for Document Image</a></h3><h3><a href="https://arxiv.org/pdf/1912.13318v5.pdf" rel="noopener noreferrer" target="_blank"> Understanding</a></h3>Presenter: <a href="https://www.linkedin.com/in/malcolm-greaves/" rel="noopener noreferrer" target="_blank">Malcolm Greaves</a> The authors try and solve the problem of automated understanding from scanned business documents. They use both computer vision and natural language processing techniques via textual and visual layout information for pre-training their system. Their system achieves SOTA on a set of layout analysis, receipt understanding, and document image classification tasks. They claim that their empirical performance is from their novel loss functions: a masked visual-language model (MVLM) loss and a multi-label document classification (MDC) loss. <img src="https://imagedelivery.net/wLbZE4_NzVVdgHc15St55g/756680db-57a2-4448-2f6f-3f0976fa5500/public" alt="Scale blog ml papers B_07"> MVLM extends the masked language model loss from BERT by randomly choosing some words to mask. They then keep the masked token’s 2D positional embedding and forces the model to predict the word’s text. MVLM forces the model to understand language context and 2D relationships between words in the document. MDC assumes that documents have possibly several labels and is a standard cross-entropy loss. This is optional and seems to help on the document image classification tasks. We found this paper interesting, but computationally demanding: alongside their loss functions, a key contribution of this paper is their pre-training routine. Without significant pre-training, these models do not seem to work well. This work requires off-the-shelf OCR and treats that as ground truth, which simplifies the work, but makes it less practical for document analysis practitioners. For the business document tasks that our ML team takes on, we often like to have greater control on the OCR task. It's sometimes advantageous to be able to adjust OCR to provide better downstream task-dependent results. We do a lot of document analysis at Scale, but can’t obfuscate or explain away the OCR component, which do produce some errors (e.g. with handwriting). Although this approach needs word-level annotations, Scale is in a unique position to get large amounts of these labor-intensive labels using our labeling platform. <h3><a href="https://arxiv.org/abs/2006.16205v1" rel="noopener noreferrer" target="_blank">Simplifying Models with Unlabeled Output Data</a></h3>Presenter: <a href="https://www.linkedin.com/in/alexandre-matton-73042b170/" rel="noopener noreferrer" target="_blank">Alexandre Matton</a> This paper focuses on tasks whose outputs have to obey some constraints. This happens in pseudo-code to code transcription for example. What we ultimately want is a code that compiles and runs without error. The usual way to address these tasks is to build a single end-to-end model and hope that it will learn the input-output mapping and understand the output constraints. The authors provide a new way to tackle these tasks by replacing the original model by two sub-models: a base predictor and a denoiser. The goal of the base predictor is to learn the mapping between the input and the output. The goal of the denoiser is to learn the constraints of the output space, in order to fix the base predictor’s output. <img src="https://imagedelivery.net/wLbZE4_NzVVdgHc15St55g/9578c3e1-2f67-408f-2bd1-a2972c25ee00/public" alt="Scale blog ml papers B_08"> There is a reason why this decomposition works well. As the denoiser simplifies the job of the base predictor, it is now possible to increase the regularization of the base predictor without affecting the overall model performance. This in turn leads to solutions that generalize better on unseen data. The big win of this architecture is that the denoiser can be trained on unlabelled output data, which is often abundant (e.g. there is a huge amount of compiling code pieces freely available on github). A big number of tasks actually impose constraints on the output (such as molecule generation, or translation from one language to another). At Scale, our ML team also tackles such problems. This paper provides a very general way to increase our downstream scores on them. <h3>Join our Team</h3> If you’re interested in joining our growing engineering team, take a look at our <a href="https://scale.com/careers" rel="noopener noreferrer" target="_blank">careers page</a> for open positions.

The machine learning research papers the Scale AI team read and discussed in Q3 2020.

Research Papers the Scale AI Team has Read in Q3 2020

Scale AI Machine Learning Digest - Q3 2020

At Scale AI, we envision a future where AI is as easy to build as any software. To help AI achieve its full potential, we are building the tools that make it easier to build and scale real-world AI systems. We began by focusing on the data - the foundation of all AI applications. With the Scale AI platform, customers can understand, visualize, and curate raw datasets, send raw data to be annotated, then debug model performance - all in one place. From the beginning, we knew the Scale AI platform had to be API-based to enable teams working on AI and ML to automate their data pipelines and seamlessly integrate Scale into ML development cycles. Since our start in 2016, our engineering team has built incredible products to deliver large volumes of training data at high-quality, whether the data input is 3D sensor data, video, image, document, or text. Despite the continued evolution of our products and the robustness of our APIs, until recently, the API documentation itself was static, with limited interactions and not a lot of details to fully understand our products, set up effective workflows, and optimize for quality.<h2>The Evolution of our Documentation:</h2> If you are a long-time Scale customer, you are probably familiar with this as our API documentation:<img src="https://imagedelivery.net/wLbZE4_NzVVdgHc15St55g/e4b49b7e-e94c-4e83-558b-ed0436a2c600/public" alt="Scale AI's Original API Documentation"> The early documentation simply served as an API reference guide. The parameters were well spaced out, included relevant examples, and showed some example code using our SDKs. As we expanded the documentation to include content on key concepts, the customer dashboard, and other elements of the Scale AI platform, the documentation became increasingly clunky and awkward to navigate. The early documentation was also very static, with limited interactions, no search capability, and overall proved to be a less than ideal user experience. <h3>Transition to ReadMe</h3> Earlier this year, our Customer Experience and Field Engineering team identified <a href="https://readme.com/" rel="noopener noreferrer" target="_blank">ReadMe</a> as an industry-leading documentation platform that could scale with our evolving needs. With ReadMe, we were able to realize three tangible benefits:<ol><li> </li><li> Unify the API reference and the Customer Dashboard documentation into a</li><li> unified website.</li><li> </li><li> </li><li> Improve the developer experience. Instead of only showing static code</li><li> examples, ReadMe enables developers to submit API requests directly with</li><li> their Scale API key. ReadMe also has built in search functionality to make</li><li> it easier for developers to find specific topics.</li><li> </li><li> </li><li> Democratize the content creation process. Instead of editing raw markdown,</li><li> making a pull request, and then deploying to a server, we just had to</li><li> update the text we needed in the ReadMe UI. Ultimately, the ReadMe</li><li> platform provided a richer and more dynamic user experience.</li><li> </li></ol> The only limitation of the ReadMe platform out-of-the-box was the lack of front-end customization to ensure visual consistency between the documentation and the rest of our website. <h3>Latest State:</h3> The latest update to our documentation marries the content and functionality of ReadMe with our front-end to provide a better and more consistent experience between our API documentation and the rest of our platform. <img src="https://imagedelivery.net/wLbZE4_NzVVdgHc15St55g/e120309c-17d5-479b-4e47-fce17cdea400/public" alt="Latest State of API Documentation"> We accomplished this by effectively leveraging ReadMe as a content management system (CMS) on the back end. We then used <a href="https://tailwindcss.com/" rel="noopener noreferrer" target="_blank">Tailwind CSS</a>, a utility-first framework to build the UI for the front end, and <a href="https://nextjs.org/" rel="noopener noreferrer" target="_blank">Next.JS</a>’s static generation feature to pre-render the page into HTML on the server ahead of each request. The HTML can be globally cached by a CDN and served instantly. Technical details on how we implemented this will follow in another blog post. <h2>New Features:</h2><h3> </h3><h3> In-depth Customer Dashboard Documentation</h3><h3> </h3> As a Scale AI customer, the Dashboard is the central location to manage your account and your projects. The new in-depth Customer Dashboard documentation walks you through how to manage your account, your projects, get an overview of submitted and completed tasks, and audit a random sample of tasks for quality. <h3>Scale 101 Workflow Overviews</h3> Every institution, company, and even team has their own ways of describing the data labeling workflow. To ensure every customer is successful on the Scale AI platform, we introduce the terminology we’ve developed in this section. <img src="https://imagedelivery.net/wLbZE4_NzVVdgHc15St55g/f3a4368d-6d0d-438c-0590-d59ddd39ad00/public" alt="Scale 101: Key Concepts"> <h3> </h3><h3> Make API calls using your Scale API key directly from the Documentation</h3><h3> </h3> Instead of static code examples, this feature enables developers to submit live API requests directly using their Scale API key. <img src="https://imagedelivery.net/wLbZE4_NzVVdgHc15St55g/8c6eac81-5b29-4e0a-0d0b-1f9feebee400/public" alt="API Key"> <h3>Search</h3> As we continue to expand the documentation, we realize it can also become more challenging to navigate to and find details on a specific topic. To offer a faster and easier way to find keywords or key phrases, we have added a new search feature in our documentation powered by Algolia. <img src="https://imagedelivery.net/wLbZE4_NzVVdgHc15St55g/f5cef952-a43c-429e-62a6-0eafe33f6e00/public" alt="Search Functionality"> <h3>Dark Mode</h3> Dark mode reduces the light emitted by device screens while maintaining the minimum color contrast ratios required for optimal readability. To view the API documentation in dark mode, select “Dark” in your computer’s display settings or “Auto” to switch from light mode to dark mode at sunset or at a specified time. <img src="https://imagedelivery.net/wLbZE4_NzVVdgHc15St55g/3c6ca586-71f7-4928-de8e-d4ee154fff00/public" alt="Light Mode/Dark Mode"> <h2>What this Means:</h2> If you are an existing Scale customer, rest assured that the updates to the documentation do not impact any existing projects, batches, tasks, or workflows. If anything, we hope these updates will clarify the expected behavior, showcase our API’s flexibility, and enable easier task creation with more effective instructions to improve the overall quality of returned tasks. If you are just getting to know our products, we hope the updated documentation inspires you to come up with new project ideas.

Introducing a new look and new features for our API Documentation

Introducing Scale AI's updated API Documentation

Introducing: Our New API Docs

At Scale AI, customers come first. This is a core credo that drives many of our daily decisions. We take the trust of our customers and the security of their data seriously, which is why we are pleased to announce the availability of our System and Organizational Controls (SOC) 2 Type 1 report. As a cloud-based data annotation provider, our customers need to be confident that we securely handle their data. SOC reports are prepared by independent Certified Public Accountant (CPA) auditors based on the internationally recognized <a href="https://www.aicpa.org/content/dam/aicpa/interestareas/frc/assuranceadvisoryservices/downloadabledocuments/trust-services-criteria.pdf" rel="noopener noreferrer" target="_blank">Trust Services Criteria</a> framework developed by the American Institute of Certified Public Accountants (<a href="https://www.aicpa.org/" rel="noopener noreferrer" target="_blank">AICPA</a>). Achieving compliance with SOC 2 requires Scale demonstrating the achievement of their service commitments and system requirements based on the Security Trust Services Category. We take our commitments to security and the responsibilities they entail seriously, which is why we have partnered with CPA firm Schellman &amp; Company, LLC to provide ourselves and, most importantly, our customers with independent, third-party validation of our processes and controls to keep our systems and customer data safe. Our commitment to enterprise-grade security is one reason leading ML teams such as NVIDIA, Toyota Research Institute, Open AI, and more rely on Scale AI to accelerate the development of their AI applications. Scale continues to invest in security best practices, and to that end, we are pursuing a Type 2 report moving forward. If you’d like further details on our SOC 2 report, please contact us at security@scale.com or sales@scale.com.

At Scale AI, we take the trust of our customers and the security of their data seriously. We are pleased to announce the availability of our SOC 2 report. 

Reaffirming Our Commitment to Building Customer Trust: Scale+ SOC 2

Reaffirming Our Commitment to Building Customer Trust: Scale + SOC 2

<h2>AI/ML Research at Scale AI:</h2> Scale AI is dedicated to accelerating the development of AI applications. As part of this mission, Scale AI’s growing AI/ML Team aims to meaningfully contribute to the AI ecosystem with research and resources that can bring significant and practical benefits to our customers and the broader research community. Research and experimentation happens throughout Scale. In some cases, our contribution is in areas where the bar has already been set high. Some examples include the open-source datasets we have released with various partners such as <a href="https://pandaset.org/" rel="noopener noreferrer" target="_blank">PandaSet</a>, <a href="https://www.nuscenes.org/" rel="noopener noreferrer" target="_blank">nuScenes</a>, and <a href="https://scale.com/resources/download/nlp-dataset-balanced-conll" rel="noopener noreferrer" target="_blank">CoNLL Balanced</a>. As our AI/ML team continues to grow, we aim to contribute in entirely new spaces by publishing cutting edge research papers. In this blog post, we spotlight a research paper we recently submitted. The research paper was a collaboration with <a href="https://www.linkedin.com/in/nivedita-suresh/" rel="noopener noreferrer" target="_blank">Nivedita Suresh</a> of Arrive. <a href="https://arriveorigin.com/" rel="noopener noreferrer" target="_blank">Arrive</a> links multi-omics and AI cell characterization to discover hidden patterns in data and improve outcomes in precision medicine.<h2> </h2><h2> Conditioning Unconditional Language Models to Recover Sentences without</h2><h2> Fine-Tuning:</h2> Pretrained language models are used as encoders with widespread success for a wide variety of NLP tasks. These models learn useful representations that can be directly applied to certain tasks. Despite the success of these models on the encoder, there isn't much research out today on whether these large pretrained language models can be used as universal decoders. To be considered "universal," a decoder must have an implicit representation for any sentence, such that it can recover that sentence exactly when conditioned on its representation. Task-specific decoders are typically trained from scratch for each task. In this research, we propose a method to try and learn useful representations of sentences using large, pretrained, transformer-based decoders. Our work investigates whether a universal decoder is even possible. If we had access to such a decoder, we could use it for a wide variety of sequence generation tasks without having to retrain or rebuild the decoder on new data. This could open up a new paradigm for low-resource machine translation, summarization, image captioning, and dialog generation.<h2> </h2><h2> The Sentence Space: How Language Models Represent Sentences</h2> Transformer-based language models like ELMo, BERT, and GPT-2 have replaced word vectors as off-the-shelf general-purpose encoders for natural language understanding. These models are trained on large amounts of text data and represent a sentence as a sequence of hidden states that come from a final layer of the transformer model. Representations in this sentence space are sequence length dependent, making comparisons between sentences with differing lengths inequitable and measuring the efficacy of using an unconditional language model as a universal decoder impossible. To resolve these issues and to make analysis easier, we propose to reparametrize the original sentence space into a lower-dimensional and sentence length agnostic vector space. We do this by adding a bias term to the fixed language model and finding the representation that minimizes the cross entropy loss of the sentence. This reparameterization now gives us the ability to project sentence tokens to representations (sentence encoding) and recover sentences from the representation (sentence recovery) via the fixed language model. We develop three representation injection mechanisms and inject biases at three different locations.<img src="https://cdn.builder.io/api/v1/image/assets%2Fe0438815ba51486bbb6a202747122d4b%2F7b93f3caad104979b5aa2042a6c7a7ff"> We add a bias Z’ to the embedding, transformer layers, and before the language modeling head in GPT-2. ’SA’ refers to selfattention, ’LN’ to layer normalization, ’FFN’ to a fully-connected layer, and ’LM Head’ to the last fully-connected layer. <h2>Experiments:</h2> Since we are testing the efficacy of a large pretrained language model as a universal decoder, we measure how well a representation can recover a sentence when injected into a language model. We measure performance across sentences from 4 different genres: books, news articles, Wikipedia, and movie dialogs. We conduct controlled experiments, varying the representation injection mechanism, representation injection location, initialization, and dimensionality of the representation. <h2>Results:</h2> We find that our representations recover sentences nearly perfectly even across genres -- achieving a BLEU score of over 98. <img src="https://imagedelivery.net/wLbZE4_NzVVdgHc15St55g/696b5ccc-0d02-4fea-4baf-16390e159600/public" alt="BLEU Scores"> BLEU score performance stratified by genre for different dimensionalities of Z. In addition, we observe that the intrinsic dimension of the representation space is the model’s latent dimension, which indicates that the language model uses its entire capacity to represent sentences. <img src="https://imagedelivery.net/wLbZE4_NzVVdgHc15St55g/0e7fa636-8519-4ff8-c7ef-019293850b00/public" alt="Sentence length v. BLEU score"> Plot of sentence length vs. BLEU score on the dataset. Additionally, our interpolation experiments reveal that the representation space has some human-understandable meaning. Our learned representations seem to have some synonym awareness. In the first sentence pair example below, the word “tale” transforms to the word “story” and the word “long” transforms to “long-running” when referring to a war. In the second example, we observe some syntactic awareness at the 0.7 mixture level. The syntax of the first sentence is retained with mostly words from the second sentence. <img src="https://imagedelivery.net/wLbZE4_NzVVdgHc15St55g/6de41532-7363-4b27-5524-0bb6e8e94800/public" alt="Linear interpolations between perfectly recovered pairs."> Two linear interpolations between perfectly recovered pairs of representations. Pink indicates token overlap to the first sentence, while blue indicates token overlap to the second sentence. <h2>Broader Impact:</h2> This work informs us that we can discover meaningful, sentence-length agnostic sentence representations, hinting at the possibility of a “universal” decoder. Such a decoder would improve low-resource sequence generation task performance and allow for considerable parameter sharing in memory and data-limited environments. <h2>Conclusion:</h2> For further details on how we set up the experiment, results, and analysis, you can find the full paper on <a href="https://arxiv.org/abs/2008.09049" rel="noopener noreferrer" target="_blank">arXiv</a>. If you’re interested in joining our growing machine learning team, you can find all of the open positions on our <a href="https://scale.com/careers?location=All&amp;team=Engineering" rel="noopener noreferrer" target="_blank">careers page</a>.

Learning useful sentence representations from large transformer-based language models.

Discovering Useful Sentence Representations from Large Pre-trained Language Models

Discovering Useful Sentence Representations from Large Pre-Trained Language Models

<h2>Introduction</h2>
Machine learning is a once-in-a-generation technological shift that is creating huge value by unlocking insights from data. But algorithmic bias remains a major concern as ML becomes more widespread. Unless ML models are trained on representative data, they can develop serious biases, significantly harming underrepresented groups and leading to ineffective products. We investigated the CoNLL-2003 dataset&mdash;a standard for building algorithms that recognize named entities in text&mdash;and found that the data is highly skewed toward male names. Using Scale&rsquo;s technology we were able to systematically mitigate this bias by:
<ol>
<li>Enriching the data to illuminate the hidden biases</li>
<li>Enhancing the dataset with underrepresented examples to mitigate gender bias</li>
</ol>
A model trained on our enhanced CoNLL-2003 dataset has both less bias and better performance&mdash;demonstrating how bias can be removed without any changes to the model. We&rsquo;ve open-sourced the Named Entity Recognition annotations for the original CoNLL-2003 dataset and our augmentation of it <a href="https://scale.com/resources/download/nlp-dataset-balanced-conll" target="_blank" rel="noopener noreferrer">here</a>.
<h2>Algorithmic bias: AI's weakness</h2>
Today, thousands of engineers and researchers are building systems that teach themselves how to achieve significant breakthroughs&mdash;improving road safety with self-driving cars, curing diseases with AI-optimized treatments, and combating climate change by managing energy usage.
But the power of self-learning systems is also their weakness. Because data is the foundation of all machine learning applications, learning from imperfect data can lead to biased outputs.
The power of AI systems means their potential to cause harm if they are biased is significant. Recent protests against the police brutality that has led to the tragic deaths of George Floyd, Breonna Taylor, Philando Castile, Sandra Bland and many others are an essential reminder of the systemic inequalities in our society that AI systems must not perpetuate. But, as we know from countless examples&mdash;whether it&rsquo;s image search results <a href="http://www.mjskay.com/papers/chi_2015_gender-bias-in-image-search.pdf" target="_blank" rel="noopener noreferrer">perpetuating gender stereotypes</a>, offender management systems <a href="https://www.propublica.org/article/machine-bias-risk-assessments-in-criminal-sentencing" target="_blank" rel="noopener noreferrer">discriminating against black defendants</a>, or facial recognition systems <a href="https://www.nist.gov/news-events/news/2019/12/nist-study-evaluates-effects-race-age-sex-face-recognition-software" target="_blank" rel="noopener noreferrer">misidentifying people of color</a> &ndash; we have a long way to go before we can say the problem of AI bias has been solved.
Bias is prevalent because it is so easily introduced. It creeps up, for example, in gold-standard open-sourced models and datasets that are the foundation of huge volumes of work in ML. The word2vec sentiment dataset for building other language models is <a href="https://blog.conceptnet.io/posts/2017/how-to-make-a-racist-ai-without-really-trying/" target="_blank" rel="noopener noreferrer">skewed by ethnicity</a>, and word embeddings&mdash;the way an ML algorithm represents words and their meanings&mdash;have been found to contain <a href="https://arxiv.org/pdf/1608.07187.pdf" target="_blank" rel="noopener noreferrer">strongly biased assumptions</a> about what occupations women are associated with.
The problem&mdash;and at least part of the solution&mdash;lie in data. To illustrate this, we&rsquo;ve conducted an experiment into one of the most popular datasets for building systems that can recognize named entities in text: CoNLL-2003.
<h2>What is Named Entity Recognition?</h2>
Named-Entity Recognition (NER) is one of the fundamental building blocks of natural language models&mdash;without it, online searches, information extraction, and sentiment analysis would be impossible.
At Scale AI, our mission is to accelerate the development of AI. Natural language is one of our main areas of focus. Our Scale Text offering includes NER&mdash;which involves annotating text according to a predefined list of labels. In practice, this might help major retailers analyze how their products are being discussed online, among other applications.
Many NER systems are trained and <a href="https://paperswithcode.com/sota/named-entity-recognition-ner-on-conll-2003" target="_blank" rel="noopener noreferrer">benchmarked</a> on CoNLL-2003&mdash;a dataset of roughly 20,000 sentences from Reuters news articles that are annotated with attributes such as &lsquo;PERSON&rsquo;, &lsquo;LOCATION&rsquo;, and &lsquo;ORGANIZATION&rsquo;.
We wanted to explore if the data was biased. To do this, we used our Scale AI labeling pipeline to categorize all the names in the dataset, asking whether they could be male, female, or either, inferring gender based on the traditional use of the name.
We found a dramatic difference. Male names are mentioned almost five times more than female names, and less than 2% of names were gender-neutral:
<img src="https://cdn.builder.io/api/v1/image/assets%2Fe0438815ba51486bbb6a202747122d4b%2F4a4527dd52b94b5ba47317dccb308c59">
This is because news articles, for societal reasons, tend to contain mostly male names. But an NER model trained on this data will be better at picking out male names than female. For example, search engines use NER models to help classify names in search queries to give more accurate results. But deploy a biased NER model, and the search engine will be worse at identifying names of women than of men&mdash;exactly the kind of subtle, pervasive bias that can creep into many real-world systems.
<h2>A novel experiment to reduce bias</h2>
To illustrate this, we trained an NER model to investigate how this gender bias would affect its performance. We built a name extraction algorithm to pick out PERSON labels using <a href="https://spacy.io/" target="_blank" rel="noopener noreferrer">spaCy</a>, a popular NLP library, and trained the model on a subset of the CoNLL data. When we tested the model on novel names in the test data that weren&rsquo;t present in the training data, we found that the model was 5% more likely to miss a new female name than a new male name&mdash;a significant discrepancy in performance:
<img src="https://cdn.builder.io/api/v1/image/assets%2Fe0438815ba51486bbb6a202747122d4b%2F728080fd7e5f47fc9ef84471bd1905d5">
We saw similar results when we used our model on the template &ldquo;NAME is a person&rdquo;, substituting the 100 most popular male and female names for each year of the US census&mdash;the model performs significantly worse on female names for all years of the census:
<img src="https://cdn.builder.io/api/v1/image/assets%2Fe0438815ba51486bbb6a202747122d4b%2Fc85d17115813441ab877a2796c4c0a6e">
Crucially, biased training data means errors are skewed toward underrepresented categories. This census experiment illustrates this a second way, too: the model&rsquo;s performance degrades significantly after 1997&mdash;the cut-off point of the Reuters articles in the CoNLL dataset&mdash;because the dataset is not representative of the popularity of names in the years since.
Models learn to fit the trends of the data they are trained on. They can&rsquo;t be expected to perform well on cases for which they have seen few examples.
<h2>If you're de-biasing the model, it's too late</h2>
So how do we fix this?
One way is to try and de-bias the model&mdash;perhaps by post-processing the model, or adding an objective function to mitigate bias, leaving it to our model to figure out the details.
But there are several reasons why this is not the best approach:
<ol>
<li>Fairness is a highly complex issue that we can&rsquo;t expect an algorithm to define on its own. Research has shown that training an algorithm to perform equally on all subsets of the population will not ensure fairness and will <a href="http://blog.mrtz.org/2016/09/06/approaching-fairness.html" target="_blank" rel="noopener noreferrer">cripple the model's learning</a>.</li>
<li>Adding extra objective functions can hurt model accuracy, causing a tradeoff. Better instead to keep the algorithm simpler and ensure the data is balanced&mdash;improving model performance and avoiding the tradeoff.</li>
<li>It&rsquo;s unreasonable to expect a model to perform well on cases for which it&rsquo;s seen few examples. The best way to ensure good results is to improve the diversity of the data.</li>
<li>Attempting to de-bias a model with engineering techniques is expensive and time-consuming. It&rsquo;s much cheaper and easier to train your models on unbiased data in the first place, freeing up your engineers to focus on applications.</li>
</ol>
Data is only one part of the bias problem. But it is foundational, affecting everything that comes after it. That&rsquo;s why we think it holds the key to some of the solution, providing potential systematic fixes at the source. Unless you explicitly label for protected classes, like gender or ethnicity, it is impossible to properly mitigate these classes as a source of bias.
This is counterintuitive. Surely, if you want to build a model that doesn&rsquo;t depend on sensitive characteristics, like gender, age or ethnicity, it&rsquo;s best to omit those properties from the training data so the model can&rsquo;t take them into account?
&ldquo;Fairness by ignorance&rdquo; actually makes the problem worse. ML models excel at drawing inferences across features&mdash;they don&rsquo;t stop doing this just because we haven&rsquo;t explicitly labeled those features. The biases simply remain undetected, making them harder to remove.
The only robust way to deal with the problem is to label more data to balance out the distribution of names. We used a separate ML model to identify sentences in the <a href="http://kdd.ics.uci.edu/databases/reuters21578/reuters21578.html" target="_blank" rel="noopener noreferrer">Reuters</a> and <a href="https://archive.org/details/BrownCorpus" target="_blank" rel="noopener noreferrer">Brown</a> corpora likely to contain female names, and then labeled those sentences via our NER pipeline to augment CoNLL.
The resulting dataset, which we&rsquo;ve called CoNLL-Balanced, has over 400 more female names. When we retrained our NER model on it, we found that our algorithm is no longer biased to perform worse on female names:
<img src="https://cdn.builder.io/api/v1/image/assets%2Fe0438815ba51486bbb6a202747122d4b%2Fef676477e28e42c68c40edba41ee4091">
Not only that, but the model also performed better at identifying male names, as well.
This is an impressive demonstration of the power of data. Mitigating bias at the source meant we did not have to make any adjustments to our ML model&mdash;saving engineering time. And we achieved this without any tradeoff in our model&rsquo;s performance&mdash;in fact, its performance slightly improved.
<img src="https://imagedelivery.net/wLbZE4_NzVVdgHc15St55g/5c767c25-9e21-47a5-717d-725505342600/public">
To allow the community to build on our work and mitigate gender bias in models built on CoNLL-2003, we&rsquo;ve open-sourced Scale AI&rsquo;s augmented dataset, which also includes gender information on <a href="https://scale.com/resources/download/nlp-dataset-balanced-conll" target="_blank" rel="noopener noreferrer">our website</a>.
The AI/ML community has its own diversity problems, but we are cautiously excited by these results. It suggests how we might be able to offer a technical solution to a pressing social problem&mdash;provided we tackle the issue head on, illuminating the hidden biases and improving the model&rsquo;s performance for everyone.
We&rsquo;re now exploring how we might apply this approach to another highly sensitive attribute&mdash;ethnicity&mdash;to explore whether we can create a robust framework for eliminating dataset bias that scales to other protected classes.
It also illustrates why we at Scale AI pay so much attention to data quality. Unless data is provably accurate, balanced, and unbiased, there&rsquo;s no guarantee that models built on it will be safe and accurate. And without that, we won&rsquo;t be able to build the transformational AI technologies that benefit everyone. If you are developing AI technologies and would like help in balancing your own datasets to ensure your models work for everyone, <a href="https://scale.com/sales" target="_blank" rel="noopener noreferrer">reach out to us</a>.
Acknowledgements:
The CoNLL 2003 Dataset referenced in this blog post is the Reuters-21578, Distribution 1.0 test collection, available on the project page for the original 2003 experiment: <a href="https://www.clips.uantwerpen.be/conll2003/ner" target="_blank" rel="noopener noreferrer">https://www.clips.uantwerpen.be/conll2003/ner/</a>.

Scale AI's platform can help mitigate bias before training a model.

The importance of detecting and mitigating bias from training data before even starting to build an algorithm.

If You’re De-Biasing The Model, It’s Too Late

We founded Scale to create the infrastructure needed to build AI in any industry, by anyone. We started tackling this complex problem at the root — turning raw data into high quality training data for models. In this pursuit, we spent the last four years building ML-augmented annotation products for all data types, expanding our solutions for major industries, and making significant technological strides in <a href="https://scale.com/blog/how-we-scale-machine-learning" rel="noopener noreferrer" target="_blank">scaling our use of ML</a> in doing so. But the problem of building effective, accurate, unbiased ML models still remains. To do this, aggregate metrics in ML are not good enough. Better ML starts with understanding your data in depth. To improve production ML, you need to understand their qualitative failure modes, fix them by gathering the right data, and curate diverse scenarios.<ol><li>Before training a model, ML engineers must curate and sample their</li><li> data, ensuring that they have the right data to solve a specific problem. This</li><li> process is too often very manual. For example, to teach a self-driving car</li><li> how to handle left turns, an ML team has to manually crawl through their</li><li> driving sequences to isolate examples of left turns to create a training</li><li> dataset. The data also needs to be representative of the ground</li><li> truth of the problem you’re trying to solve. If you’re building a model to</li><li> assign a gender to faces, you have to ensure representative data for all</li><li> genders to have unbiased outputs. Again, that’s too often a manual and</li><li> highly inefficient process.</li><li> </li><li>After training, ML teams test and benchmark model performance, ensuring that the test dataset is also sufficiently representative of the</li><li> problem they are trying to assess its performance against. For example, a</li><li> model learning how to tell cars apart from pedestrians needs enough of both</li><li> examples for accurate benchmarking. This requires ML engineers to spend</li><li> significant amounts of time building one-off UIs to chart and share</li><li> performance data.</li><li> </li><li>After deployment, debugging the model, identifying failure</li><li> modes and fixing them. Too often issues in data only arise after a model has</li><li> entered deployment – requiring time-consuming debugging. One Scale customer,</li><li> for example, found that their vehicle recognition algorithm didn’t perform</li><li> well in certain environments – it turned out that the model was trained on a</li><li> dataset where vehicles were mostly in the bottom of the image, so the model</li><li> associated “bottom of the image” with “likelihood of being a car.”</li><li> </li><li> </li></ol> The Scale team has been working to productize the concept that Andrej Karpathy calls "<a href="https://www.youtube.com/watch?v=oBklltKXtDE&amp;feature=youtu.be&amp;t=475s" rel="noopener noreferrer" target="_blank">Operation Vacation</a>." Nucleus is a new way—the right way—to develop ML models, helping us move away from the concept of one dataset and towards a paradigm of collections of scenarios and giving ML engineers the ability to automate time-consuming manual steps in the ML development process. <img src="https://cdn.builder.io/api/v1/image/assets%2Fe0438815ba51486bbb6a202747122d4b%2F004680ecf1424e3193faf21d626662b8" alt="ML Development Cycle"><a href="https://scale.com/nucleus" rel="noopener noreferrer" target="_blank">Scale Nucleus</a> provides advanced tooling for understanding, visualizing, curating, and collaborating on your data – allowing teams to build better ML models via a powerful interface and APIs. With Scale Nucleus, you can: <ul><li>Visualize your dataset, ground truth, and model</li><li> predictions to improve model performance</li><li> </li><li>Curate interesting slices within your dataset for active</li><li> learning and identifying key edge cases</li><li> </li><li>Upload and choose data to be annotated for rare event</li><li> mining and dataset balancing</li><li> </li><li>Search your data based on metadata or ML-produced</li><li> attributes</li><li> </li><li>Identify edge cases through visual search</li><li>Measure key metrics like dataset balance, class</li><li> correlation, and confusion, via a powerful insights tab that shows the</li><li> overall health of the data</li><li> </li><li>Debug model performance</li><li>Share your dataset seamlessly to provide a single source</li><li> of truth for data within your team.</li><li> </li></ul> <img src="https://imagedelivery.net/wLbZE4_NzVVdgHc15St55g/a41b6e07-5da7-41e2-5d94-542a8a933900/public" alt="Scale Nucleus Query"> Scale Nucleus’s query function, allowing users to return images matching automatically-generated meta-tags. Scale Nucleus is directly integrated with Scale AI’s data labeling pipeline – allowing teams to fix any issues in their data at the source. Nucleus currently supports Image data, with support for 3D Sensor Fusion, Video, Text and Document data coming soon. We built Scale Nucleus to allow data scientists and ML engineers to manage data more efficiently and increase the marginal value of their data. Deeper insight into a dataset’s features can have several transformational effects on the development of AI: <ul><li> </li><li> For engineers, the easier visualization and curation of datasets lowers</li><li> the barriers to entry to building ML systems.</li><li> </li><li> </li><li> Spotting hidden modes of failure in datasets before deployment (such as a</li><li> set of driving sequences that doesn’t include any sequences at night)</li><li> makes it much easier to train high-quality models, and provides a robust</li><li> way of eliminating issues like bias at source.</li><li> </li><li> </li><li> The ease of debugging could significantly improve iteration speed on model</li><li> fine-tuning post-deployment.</li><li> </li></ul> We are excited to take the next step in building the infrastructure to enable efficient, accurate, and unbiased ML development. If you’d like to join us in this journey and try Scale Nucleus, contact us at nucleus@scale.com or sign up on our <a href="https://scale.com/nucleus#sales" rel="noopener noreferrer" target="_blank">website</a>.

Scale Nucleus: The mission control for your data

Introducing Scale AI's latest product: Scale Nucleus

Introducing: Scale Nucleus

URL redirections for Website. Disclaimer: takes 5~ min for each added or edited.

url-redirects

Use relative path. use `/blog/foo` not `https://scale.com/blog/foo`

blog-filter-type

homepage-quotes

logos

homepage-content

quotes

blog-category-model

announcement-banner

homepage-resources

homepage-event

custom-fonts

author

Event model for new marketing events page

event

Optional second image intended to show on the grid view, should be a more squared image closer to 1:1 aspect ratio than the one used in the Cover field.

guides

announcement-news

homepage-research

legal

slugs must be all lowercase. ie: ‘my-slug’

eval

datastream

customers

Display order on the page /customers. Higher values show first

summit-2023-speakers

event-model

Toggle if the event is featured. Featured events will be moved to the top of the list

blog-model

Option to list or not list this post in the Blog index. Enabled = not list in the index, Disabled = list in the index

Products

Enterprise

Government

Resources