Apr
21
2020
--

AWS and Facebook launch an open-source model server for PyTorch

AWS and Facebook today announced two new open-source projects around PyTorch, the popular open-source machine learning framework. The first of these is TorchServe, a model-serving framework for PyTorch that will make it easier for developers to put their models into production. The other is TorchElastic, a library that makes it easier for developers to build fault-tolerant training jobs on Kubernetes clusters, including AWS’s EC2 spot instances and Elastic Kubernetes Service.

In many ways, the two companies are taking what they have learned from running their own machine learning systems at scale and are putting this into the project. For AWS, that’s mostly SageMaker, the company’s machine learning platform, but as Bratin Saha, AWS VP and GM for Machine Learning Services, told me, the work on PyTorch was mostly motivated by requests from the community. And while there are obviously other model servers like TensorFlow Serving and the Multi Model Server available today, Saha argues that it would be hard to optimize those for PyTorch.

“If we tried to take some other model server, we would not be able to quote optimize it as much, as well as create it within the nuances of how PyTorch developers like to see this,” he said. AWS has lots of experience in running its own model servers for SageMaker that can handle multiple frameworks, but the community was asking for a model server that was tailored toward how they work. That also meant adapting the server’s API to what PyTorch developers expect from their framework of choice, for example.

As Saha told me, the server that AWS and Facebook are now launching as open source is similar to what AWS is using internally. “It’s quite close,” he said. “We actually started with what we had internally for one of our model servers and then put it out to the community, worked closely with Facebook, to iterate and get feedback — and then modified it so it’s quite close.”

Bill Jia, Facebook’s VP of AI Infrastructure, also told me, he’s very happy about how his team and the community has pushed PyTorch forward in recent years. “If you look at the entire industry community — a large number of researchers and enterprise users are using AWS,” he said. “And then we figured out if we can collaborate with AWS and push PyTorch together, then Facebook and AWS can get a lot of benefits, but more so, all the users can get a lot of benefits from PyTorch. That’s our reason for why we wanted to collaborate with AWS.”

As for TorchElastic, the focus here is on allowing developers to create training systems that can work on large distributed Kubernetes clusters where you might want to use cheaper spot instances. Those are preemptible, though, so your system has to be able to handle that, while traditionally, machine learning training frameworks often expect a system where the number of instances stays the same throughout the process. That, too, is something AWS originally built for SageMaker. There, it’s fully managed by AWS, though, so developers never have to think about it. For developers who want more control over their dynamic training systems or to stay very close to the metal, TorchElastic now allows them to recreate this experience on their own Kubernetes clusters.

AWS has a bit of a reputation when it comes to open source and its engagement with the open-source community. In this case, though, it’s nice to see AWS lead the way to bring some of its own work on building model servers, for example, to the PyTorch community. In the machine learning ecosystem, that’s very much expected, and Saha stressed that AWS has long engaged with the community as one of the main contributors to MXNet and through its contributions to projects like Jupyter, TensorFlow and libraries like NumPy.

Mar
16
2020
--

Torch & Everwise merge into affordable exec coaching for all

While companies might pay for a CEO coach, lower level employees often get stuck with lame skill-building worksheets or no mentorship at all. Not only does that limit their potential productivity, but it also makes them feel stagnated and undervalued, leading them to jump ship.

Therapy… err… executive coaching is finally becoming destigmatized as entrepreneurs and their teams realize that everyone can’t be crushing it all the time. Building a business is hard. It’s okay to cry sometimes. But the best thing you can do is be vulnerable and seek help.

Torch emerged from stealth last year with $18 million in funding to teach empathy to founders and C-suite execs. Since 2013, Everwise has raised $26 million from Sequoia and others for its peer-to-peer mentorship marketplace that makes workplace guidance accessible to rank-and-file staffers. Tomorrow they’ll official announce their merger under the Torch name to become a full-stack career coach for every level of employee.

“As human beings, we face huge existential challenges in the form of pandemics, climate change, the threats coming down the pipe from automation and AI” says Torch co-founder and CEO Cameron Yarbrough. “We need to create leaders at every single level of an organization and ignite these people with tools and human support in order to level up in the world.”

Startup acquisitions and mergers can often be train wrecks because companies with different values but overlapping products are jammed together. But apparently it’s gone quite smoothly since the products are so complementary, with all 70 employees across the two companies keeping their jobs. “Everwise is much more bottom up whereas Torch is about the upper levels, and it just sort of made sense” says Garry Tan, partner and co-founder of Initialized Capital that funded Torch’s Series A and is also a client of its coaching.

How does each work? Torch goes deep, conducting extensive 360-interviews with an executive as well as their reports, employees, and peers to assess their empathy, communication, vision, conflict resolution, and collaboration. Clients’ executives do extensive 360-interviews. It establishes quantifiable goals that executives work towards through video call sessions with Torch’s coaches. They learn about setting healthy workplace boundaries, staying calm amidst arguments, motivating staff without seeming preachy, and managing their own ego.

This coaching can be exceedingly valuable for the leaders setting a company’s strategy and tone. But the one-on-one sessions are typically too expensive to buy for all levels of employees. That’s where Everwise comes in.

Everwise goes wide, offering a marketplace with 6,000 mentors across different job levels and roles that can provide more affordable personal guidance or group sessions with 10 employees all learning from each other. It also provides a mentorship platform where bigger companies can let their more senior staffers teach junior employees exactly what it takes to succeed. That’s all stitched together with a curated and personalized curriculum of online learning materials. Meanwhile, a company’s HR team can track everyone’s progress and performance through its Academy Builder dashboard.

“We know Gen Z has grown up with mentors by their side from SAT prep” says Torch CMO Cari Jacobs. Everwise lets them stay mentored, even at early stages of their professional life. “As they advance through their career, they might notch up to more executive private coaching.” Post-merger, Torch can keep them sane and ambitious throughout the journey. 

“It really allows us to move up market without sacrificing all the traction we’ve built working with startups and mid-market companies,” Yarbrough tells me. Clients have included Reddit and ZenDesk, but also giants like Best Buy, Genentech, and T-Mobile.

The question is whether Everwise’s materials are engaging enough to not become just another employee handbook buried on an HR site that no one ever reads. Otherwise, it could just feel like bloat tacked onto Torch. Meanwhile, scaling up to bigger clients pits Torch against long-standing pillars of the executive coaching industry like Aon and Korn Ferry that have been around for decades and have billions in revenue. Meanwhile, new mental health and coaching platforms are emerging like BetterUp and Sounding Board.

But the market is massive since so few people get great coaching right now. “No one goes to work and is like, ‘Man, I wish my boss was less mindful,’” Tan jokes. When Yarbrough was his coach, the Torch CEO taught the investor that while many startup employees might think they thrive on flexibility, “people really want high love and high structure.” In essence, that’s what Torch is trying to deliver — a sense of emotional camaraderie mixed with a prod in the direction of fulfilling their destiny.

Powered by WordPress | Theme: Aeros 2.0 by TheBuckmaker.com