/* ---- Google Analytics Code Below */
Showing posts with label Open Source. Show all posts
Showing posts with label Open Source. Show all posts

Sunday, July 23, 2023

Meta launches Llama 2 open-source LLM

The march continues.

Meta launches Llama 2 open-source LLM

About the AuthorBy Ryan Daws | July 19, 2023

Categories: Companies, Development, Machine Learning, Meta (Facebook),

Meta launches Llama 2 open-source LLMRyan is a senior editor at TechForge Media with over a decade of experience covering the latest technology and interviewing leading industry figures. He can often be sighted at tech conferences with a strong coffee in one hand and a laptop in the other. If it's geeky, he’s probably into it. Find him on Twitter (@Gadget_Ry) or Mastodon (@gadgetry@techhub.social)

Meta has introduced Llama 2, an open-source family of AI language models which comes with a license allowing integration into commercial products.

The Llama 2 models range in size from 7-70 billion parameters, making them a formidable force in the AI landscape.

According to Meta’s claims, these models “outperform open source chat models on most benchmarks we tested.”

The release of Llama 2 marks a turning point in the LLM (large language model) market and has already caught the attention of industry experts and enthusiasts alike.

The new language models offered by Llama 2 come in two variants – pretrained and fine-tuned:

The pretrained models are trained on a whopping two trillion tokens and have a context window of 4,096 tokens, enabling them to process vast amounts of content at once.

The fine-tuned models, designed for chat applications like ChatGPT, have been trained on “over one million human annotations,” further enhancing their language processing capabilities.

While Llama 2’s performance may not yet rival OpenAI’s GPT-4, it shows remarkable promise for an open-source model.

The long-awaited sequel, Llama-2 is announced today! It's the best OSS model we have now.  ... ' 

Saturday, July 08, 2023

China Releases Its First Open Source Operating System

 China Releases Its First Open Source Operating System

Reuters, Josh Ye, July 6, 2023

Chinese state media reports that nation has released its first domestic open source desktop computer operating system, in a bid to reduce its dependence on U.S. technology. About 4,000 developers built the Linux-based OpenKylin system, which China's space program, and industries including finance and energy, already are using. China's technology industry has focused on designing an operating system that has no reliance on U.S. tech in recent years; more than a dozen Chinese companies are involved in that effort.  ... ' 

Friday, December 16, 2022

Overture Maps Emerge to Compete with Google

 Nice to see Competitive Geographical mapping, been using Google for some time.   Google has done a great job,  but want to see yet more creative things.   Open Source I see.   Note the number of participants. 

Linux, Amazon, Meta, and Microsoft want to break the Google Maps monopoly

Overture Maps Foundation wants to end the oppressive rule of the Google Maps API.   By Ron Amadeo - 12/16/2022, 

Google Maps is getting some competition. The Linux Foundation has announced Overture Maps, a "new collaborative effort to develop interoperable open map data as a shared asset that can strengthen mapping services worldwide." It's an open source mapping effort that includes a list of heavy hitters: Amazon Web Services (AWS), Meta, Microsoft, and TomTom, with the foundation adding that the project is "open to all communities with a common interest in building open map data."

The Linux Foundation has a press release about the project and a new website for the Overture Maps Foundation. The press release outlined the scope of the project, which aims to deliver:

Collaborative Map Building: Overture aims to incorporate data from multiple sources including Overture Members, civic organizations, and open data sources.

Global Entity Reference System: Overture will simplify interoperability with a system that links entities from different data sets to the same real-world entities.

Quality Assurance Processes: Overture data will undergo validation to detect map errors, breakage, and vandalism to help ensure that map data can be used in production systems.

Structured Data Schema: Overture will define and drive adoption of a common, structured, and documented data schema to create an easy-to-use ecosystem of map data.

If you're saying, "Wait! isn't there already an open source map community out there?"  There is, and it's called "OpenStreetMap," the Wikipedia of maps that anyone can edit. The Overture press release says, "The project will seek to integrate with existing open map data from projects such as OpenStreetMap and city planning departments, along with new map data contributed by members and built using computer vision and AI/ML techniques to create a living digital record of the physical world."

Monday, October 31, 2022

IBM and RedHat

Following this for some time.  Here Marketwatch calls it a big win. 

Opinion: IBM’s expensive and risky pivot to the cloud is finally paying off

Published: Oct. 25, 2022 at 3:10 p.m. ET,  By Alex Salkever

Red Hat is a bright spot for the company, which is starting its third transformation in history

International Business Machines agreed to buy open-source enterprise-software company Red Hat for a whopping $33 billion in 2019, its biggest acquisition ever.

Critics scoffed that the deal was folly. Everyone knew it was a big bet, even for a behemoth such as IBM IBM, +2.78%. The purchase also represented a one-way door that IBM chose to go through — a shift to become primarily an open-source software company.

Red Hat’s business was built on selling services and support for software that anyone could freely download and deploy. Yes, the rising popularity of open-source meant Red Hat would likely continue to be a critical component in enterprise computing.

Read: This one stock stands out when looking at key data from the FAANG+ group

But those critics thought Big Blue would struggle to be one of the open-source cool kids and that its old-school culture would kill Red Hat’s vibe. Most importantly, Red Hat’s business model could never hope to touch the profitability of Big Blue’s tidy proprietary software and hardware businesses. .... ' 

(Much More) 

Wednesday, February 09, 2022

Open Source Lab Tools

Had not thought of this, clever idea as it is applied.

Bringing Open Source to the Global Lab Bench

BY JULIETA ARANCIO, SHANNON DOSEMAGEN   in Issues.org

In 2015, Richard Bowman, an optics scientist, began experimenting with 3D printing a microscope as a single piece in order to reduce the time and effort of reproducing the design. Soon after, he started the OpenFlexure project, an open-license 3D-printed microscope. The project quickly took over his research agenda and grew into a global community of hundreds of users and developers, including professional scientists, hobbyists, community scientists, clinical researchers, and teachers. Anyone with access to a 3D printer can download open-source files from the internet to create microscopes that can be used for doing soil science research, detecting diseases such as malaria, or teaching microbiology, among other things. Today, the project is supported by a core team at the Universities of Bath and Cambridge in the United Kingdom, as well as in Tanzania by the Ifakara Health Institute and Bongo Tech & Research Labs, an engineering company. 

OpenFlexure is one of many open science hardware projects that are championed by the Gathering for Open Science Hardware (GOSH), a transnational network of open science hardware advocates. Although there are differences in practice, open hardware projects operate on similar principles to open-source software, and they span disciplines ranging from nanotechnology to environmental monitoring. GOSH defines the field as “any piece of hardware used for scientific investigations that can be obtained, assembled, used, studied, modified, shared, and sold by anyone. It includes standard lab equipment as well as auxiliary materials, such as sensors, biological reagents, analog and digital electronic components.” Compared to an off-the-shelf microscope, which may cost thousands of dollars, an OpenFlexure microscope may cost a few hundred. By being significantly cheaper and easier to maintain, open hardware enables more people in more places to do science.

The academic production of open hardware has increased exponentially in the last five years with the emergence of networks, dedicated publication venues, peer-reviewed literature, and thematic events, all aimed at supporting open science hardware projects. Successful projects have led to greater access to equipment, which increases research efficiency, and have fostered equity and public participation in science. However, open hardware is still the exception in science, and the designs of most research tools remain unavailable to their users, which limits their accessibility and adaptability. Encouraging broader use of open science hardware will require funding agencies, universities, and international organizations to cooperate and incentivize researchers to develop and share hardware designs.

OPEN HARDWARE SOLVES AN ACCESS PROBLEM AND PRODUCES BETTER SCIENCE TOOLS 

Open hardware addresses a significant hurdle for global scientists: science equipment is often expensive to purchase and difficult or impossible to customize or repair. New research questions—or questions in new research settings—often require that tools be modified or customized. Lack of access to designs makes tools more difficult to customize, leading to delays and additional costs.

Open hardware is still the exception in science, and the designs of most research tools remain unavailable to their users, which limits their accessibility and adaptability.

Historically, science hardware manufacturing has been concentrated in well-resourced countries, making it unreasonably difficult and expensive to obtain and repair elsewhere. As a result, researchers, educators, and others outside academia and at the periphery of these production centers find it almost impossible to access and maintain research infrastructure. The World Health Organization estimates that 70% of donated medical equipment in sub-Saharan Africa is out of service or not in use due to lack of authorized service engineers and necessary parts.   .... ' 

Sunday, October 10, 2021

Google Pledges a Million to Secure Open Source

Securing Open Source.

Google Pledges $1 million to Secure Open Source Software

By Computing (U.K.) October 5, 2021

Google has pledged $1 million in funding to a new open source security project hosted by the Linux Foundation.

Dubbed the Secure Open Source (SOS) Rewards program, the pilot project aims to improve the security of critical open-source apps by offering eligible developers financial rewards of $10,000 or more for their security-related work.

"SOS rewards a very broad range of improvements that proactively harden critical open source projects and supporting infrastructure against application and supply chain attacks," Meder Kydyraliev and Kim Lewandowski of Google's Open Source Security Team said in a blog post.

"To complement existing programs that reward vulnerability management, SOS's scope is comparatively wider in the type of work it rewards, in order to support project developers."

From Computing (U.K.)

Tuesday, June 15, 2021

IBM Open Sources Hyperledger

 Had examined this sometime ago, seems IBM wants it and blockchain to succeed.  This will promote it it be be better tested. 

IBM Donates Code Improvements to Open Source Hyperledger

“The intent is to improve the usability of Hyperledger for all users,” said Christopher Ferris, CTO at IBM.

By Tanzeel Akhtar in Coindesk

 (NYSE: IBM) said it has contributed significant code to the enterprise blockchain network Hyperledger, “one of the largest in its history.”

IBM announced Wednesday it has donated updates to the Hyperledger code and the underlying code for the IBM Blockchain Platform Console, an intuitive user interface that will form what is now the Fabric Operations Console.

The firm said it is also donating the code that supports token exchanges on Hyperledger Fabric and the Fabric Token SDK, as well as support for interoperability among different distributed ledger technology (DLT) platforms.  .... " 

Friday, February 12, 2021

Standardizing Robot Operating Systems

Just because they use the same software does not mean they will 'get along'.   Still the use of software that can communicate will make them easier to update and exchange information.  

As Robots Fill the Workplace, They Must Learn to Get Along  By Wired, February 12, 2021

Singapore's Changi General Hospital is using software from Open Robotics to avoid standoffs among its robots in hallways or outside elevators.

The hospital's approximately 50 robots are from eight manufacturers, and the software enables robots from different manufacturers to communicate and negotiate safe passage around each other.  Open Robotics maintains the Robot Operating System (ROS) open-source software, but the software being used by Changi allows communication among robots not based on ROS.

The company hopes for more widespread adoption of such free, easily modified software to facilitate greater interoperability of workplace robots.

ABI Research's Rian Whitton said companies have been able to develop mobile robots more quickly thanks to ROS, and interoperability could accelerate through an open source platform.

From Wired  ... 

Friday, January 08, 2021

Data Lineage Tools

This problem was brought to my attention.     I don't buy that these are the best,  but they are reasonable examples I will examine.   Especially with regard to impact and potential risk to the outcome of business process.   We built our own then,  prebuilt here.  It is a key part of the intelligence behind data and its use. 

 " .... Data lineage tools document the flow of data in and out of organization systems. They capture end-to-end lineage and ensure impact and risk analysis can be performed in the event of problems or changes to data assets as they move across pipelines.  ... "

The 8 Best Open-Source Data Lineage Tools to Consider   by 7wData

Searching for data integration and data management software can be a daunting (and expensive) process, one that requires long hours of research and deep pockets. The most popular enterprise data lineage tools often provide more than what’s necessary for non-enterprise organizations, with advanced functionality relevant to only the most technically savvy users. Thankfully, there are a distinct group of the best open-source data lineage tools out there. Some of these solutions are offered by vendors looking to eventually sell you on their enterprise product, and others are maintained and operated by a community of developers looking to democratize the process.

In this article, we will examine the best open-source data lineage tools, first by providing a brief overview of what to expect and also with short blurbs about each of the currently available options in the space. This is the most complete and up-to-date directory on the web.

Apatar

Apatar is a free and open-source data integration software package designed to help business users and developers move data in and out of a variety of data sources and formats. The tool requires no programming or design to accomplish even complex integration with joins across several data sources. Apatar provides a visual interface to minimize the impact of system changes. The tool comes with a pre-built set of integration tools and enables users to re-use previously built mapping schemas as well.

CloverETL

CloverETL (now CloverDX) was one of the first open-source ETL tools. The Java-based data integration framework was designed to transform, map, and manipulate data in various formats. CloverETL can be used standalone or embedded and connects to RDBMS, JMS, SOAP, LDAP, S3, HTTP, FTP, ZIP, and TAR. Though the product is no longer offered by the provider, it can be downloaded securely using SourceForge . CloverDX still supports CloverETL in line with their standard support agreement as well.   

Dremio

Dremio offers a product called Data Lake engine that provides fast query speed and a self-service semantic layer that operates directly against Data Lake storage. The solution connects to S3, ADLS, Hadoop or wherever enterprise data resides. Apache Arrow, Data Reflections and other Dremio technologies work together to hasten query speeds, and the semantic layer enables IT to apply security and business meaning. Users do not have to send data to Dremio or have it stored in proprietary formats to access it. .... (more below) .... 

Sunday, August 23, 2020

Connected Cars Run on Open Source

Why and how this is being put together as a broader solution.

Why the connected car rides on open source in VentureBeat
Tom Canning
August 23, 2020 12:12 PM
Transportation

The automobile is one of the most exciting frontiers in our connected lives today. Infotainment systems, real-time maps, and advanced driver assistance systems are already commonplace in newer vehicles, but it is still relatively early days for the fully connected car. The exhilarating vision of the driving experience of the near future includes augmented reality dashboards, progressively more autonomous operations, and increased integration with the outside world, from interacting with smart home devices to automatically finding parking spots nearby.   ... " 

Saturday, July 11, 2020

A Problem with Google's Open Usage Commons

IBM Suggests there is a Problem:

IBM has a problem with Google’s Open Usage Commons
The ties between the Open Usage Commons and Google may be too strong and clear.  By Jim Salterin in ArsTechnia

This Wednesday, Google announced a new open source initiative—the Open Usage Commons, a sort of stewardship project for open source trademarks. The move drew immediate criticism from IBM, which claims an interest in Istio, one of the three projects Google seeded the OUC with at launch.  ... "

From the IBM DeveloperBlog.

Friday, July 10, 2020

A New Kind of Open Source: Open Commons

Google Open Usage Commons.   Considering this, there will be more to follow.  Thoughts?

Announcing a new kind of open source organization  Google Blog.
Wednesday, July 8, 2020

Google has deep roots in open source. We're proud of our 20 years of contributions and community collaboration. The scale and tenure of Google’s open source participation has taught us what works well, what doesn’t, and where the corner cases are that challenge projects.

One of the places we’ve historically seen projects stumble is in managing their trademarks—their project’s name and logo. How project trademarks are used is different from how their code is used, as trademarks are a method of quality assurance. This includes the assurance that the code in question has an open source license. When trademarks are properly managed, project maintainers can define their identity, provide assurances to downstream users of the quality of their offering, and give others in the community certainty about the free and fair use of the brand.

In collaboration with academic leaders, independent contributors, and SADA Systems, today we are announcing the Open Usage Commons, an organization focused on extending the philosophy and definition of open source to project trademarks. The mission of the Open Usage Commons is to help open source projects assert and manage their project identity through programs specific to trademark management and conformance testing. Creating a neutral, independent ownership for these trademarks gives contributors and consumers peace of mind regarding their use of project names in a fair and transparent way.

Understanding and managing trademarks is critical for the long-term sustainability of projects, particularly with the increasing number of enterprise products based on open source. Trademarks sit at the juncture of the rule of law and the philosophy of open source, a complicated space; for this reason, we consider it to be the next challenge for open source, one we want to help with.

To get the Open Usage Commons started, Google has contributed initial funding, and the trademarks of Angular, a web application framework for mobile and desktop; Gerrit, web-based team code-collaboration tool; and Istio, an open platform to connect, manage, and secure microservices, will be joining the Open Usage Commons. If you use a trademark of one of the projects currently, you can continue to use those marks, following any current guidance from the project. As the Open Usage Commons is focused on trademark management, the contributor communities and technical roadmaps of these projects are not changed by joining the Commons, although we hope this new model encourages anyone who has stood on the sidelines until now to participate in these projects .... " 

Wednesday, June 10, 2020

Clarity Language for Smart Contracts

Brought to my attention as a part of a continuing look at smart contracts for supply chain and related analytic systems.

Clarity Language for Smart Contracts

Clarity Language
Clarity is a new language for smart contracts. Clarity is a decidable language, meaning you can know, with certainty, from the code itself what the program will do. Clarity is interpreted (not compiled) and the source code is published on the blockchain. Clarity gives developers a safe way to build complex smart contracts. The Clarity open-source project is supported by Blockstack and Algorand. .... 

The future of Smart Contracts
Clarity recognizes the need for smart contract languages that are more safe, secure, and predictable. Our goal is to help the smart contract industry mature beyond its current state. Clarity is designed by scientists from Princeton and MIT, and the open-source project is initially supported by Blockstack and Algorand. We welcome other projects and developers to join this effort and help contribute! .... '

In Github: https://github.com/clarity-lang  .... 

Algorand and Blockstack join to adopt Clarity blockchain smart contract language  By Kyt Dotson in SIliconAngle

Two blockchain industry leaders, Algorand Inc. and Blockstack PBC, today announced an independent open-source project to support Clarity, the blockchain distributed ledger smart contract language.

A smart contract is a programmable agreement that automates actions based on the completion of agreed-upon terms between multiple parties that have been encoded. The code of the smart contract and the agreements made exist within a blockchain, which acts as a tamperproof record of the contract, as well as assets under contract, which could be money, property, shares or anything else of value that can be represented digitally.   .... '

Thursday, March 26, 2020

Wal-Mart Joins Hyperledger

Wal-Mart's early looks at this seem to be identification and tracking related supply chain applications.    Is this an indication of further dives into the space?

Walmart Joins Hyperledger Alongside 7 Other Companies  By Samuel Haig

Walmart has become the latest major conglomerate to join open-source blockchain consortium Hyperledger. Walmart is among eight new members to join the platform. The new members were announced on March 3 at the Hyperledger Global Forum 2020 in Phoenix, Arizona.

Sanjay Radhakrishnan, the vice president of Walmart Global Tech, expressed excitement in joining the platform, stating:

“We've seen strong results through our various deployments of blockchain, and believe staying involved in open source communities will further transform the future of our business."  ... '

Wednesday, March 25, 2020

Personality Key for Open Source Contribution?

Depends if we can really detect useful personality measures of this type, and if they will be stable under differing contexts and goals.

Personality Key in Whether Developers Can Contribute to Open Source Projects
Waterloo News

The results of a study by researchers at the University of Waterloo in Canada suggest a software developer's personality could affect their ability to contribute to open source projects. Although social factors are the primary determinant of acceptance or rejection of online contributors' work, Waterloo's Meiyappan Nagappan said personality also is important to consider because it governs how contributors' behaviors manifest in their interactions with others. The researchers assessed data from the GitHub open source platform to analyze the personality traits of 16,935 active developers from 1,860 projects, and extracted the five leading developer personalities—openness, conscientiousness, extraversion, agreeableness, and neuroticism—with the IBM Watson Personality Insights service. Waterloo's Alex Yun said the analysis suggested that biases may be involved in the acceptance or rejection of contributions to work on open source platforms. Said Yun, "Managers are more likely to accept a contribution from someone they know, or someone more agreeable than others, even though the technical contribution might be similar."

Monday, January 13, 2020

Sharing with a Core Scientific Data Model

With data getting more complex, and havig to include many aspects of context-specific metadata, sharing it efficiently can be messy.  In our experience the integration of needed metadata can be problematic.

Simplifying How Scientists Share Data    By Ohio State News

Their system is designed to share complex datasets easily, without massive files, and without losing metadata.   Researchers have created a new open-source data-management system for scientists, in the hope the system might make collaboration easier.

Researchers at The Ohio State University, Denmark's University of Aarhus, and France's University of Orléans have developed an open-source data-management system that will help scientists share data with each other.

The system, called the "Core Scientific Data Model," is designed to share complex datasets easily, without massive files that take up a lot of bandwidth and hard drive space, and without losing metadata.

The researchers built the software, uploaded it to the web, and made the code open-source, with the goal of creating a simple, free way to combine multiple types of data into one place.

Said Ohio State researcher Deepansh Srivastava, "Instead of looking for data and plucking it from datasets, if we could simply export it as this one file type – as a core scientific data file type – we'd be able to work in a common system."

From Ohio State News

Friday, August 09, 2019

Will MyCroft Compete with Best AI Assistants?

Had seen Mtcroef some time ago, in 2015, then also touting its Open Source.  Still has a long way to go to compete.   Will try to get a demo of it in context.  I think its unlikely the privacy issues will make a difference as they become patched.

Can an open-source AI take on Amazon and Google?
Mycroft certainly hopes so.

By Nicole Lee, @nicole in Engadget

It's only been a few years since Amazon unveiled the Alexa-powered Echo, but since then, smart speakers have become a major consumer-electronics category. Key to its success is the notion of the always-on virtual assistant, which other companies like Apple and Google have adopted as well. In fact, not only has Google made Assistant the driving force behind its Android smartphones, it has launched its own line of Echo rivals.

But underneath all of this technology is the potential risk to your privacy. In just the past few months, news reports have uncovered a series of alarming revelations that companies like Amazon, Google and even Apple have been listening in on conversations without permission. The data that they collect are also often stored indefinitely unless you explicitly delete it or turn off the recording ability. The companies have since responded that the listening of information only occurs to a small percentage of its customers and that the data are anonymized. While that may be true, it's disconcerting that none of this is transparent, and that the customer is rarely in complete control of their data.

An easy way to avoid this, of course, is to not partake in this technology at all. But a company based in Lawrence, Kansas, is working on an alternative solution: a virtual assistant aimed at preserving privacy and that's also open source. It's called Mycroft, and though you may not have heard of it, the company's been around since 2015.

"At the time, the only real voice technology that was in broad production was Siri," said Mycroft CEO and founder Joshua Montgomery to Engadget. Amazon had announced Alexa in November of 2014, but it was still in private beta when the idea for Mycroft came about. "We thought, you know, hey, this is the type of technology that could be really groundbreaking in the future. And scarcely a year later [in 2016], Amazon launched their Super Bowl ad, and then Google got in on the game, and suddenly it's the fastest-growing segment of the technology market."  .... "

Saturday, March 23, 2019

AI and Patents in the WIPO

In a meeting a few weeks ago I asked the question about how findings using AI methods were being protected,  and got a somewhat unclear answer.    This perhaps shows why. Here much more about what is happening in this space.  Interesting stats about directions. - Franz

Based on the:WIPO (World Intellectual Property Organization) Trends Report  (158 page PDF) 

AI and patents - trends to watch  from the WIPO technology trends report
WIPO Technology Trends report 2019  came as a surprise to me. We in AI are not used to thinking about patents so much because tools / platforms are mostly Open sourced.

Here are the key takeaways from this report:
AI and patents - trends to watch  from the WIPO technology trends report.Posted by Ajit Jaokar in the DSC

The ratio of scientific papers to inventions has decreased from 8:1 in 2010 to 3:1 in 2016 – indicative of a shift from theoretical research to the use of AI technologies in commercial products and services.
Among AI functional applications, computer vision, which includes image recognition, is the most popular. Computer vision is mentioned in 49 percent of all AI-related patents (167,038 patent documents), growing annually by an average of 24 percent (21,011 patent applications filed in 2016).  

functional applications with the highest growth rates in patent filings in the period 2013 to 2016 were AI for robotics and control methods, which both grew on average by 55 percent a year.  
Many AI patents include inventions that can be applied in different industries – telecommunications, transportation , medical sciences , personal devices, computing and human–computer interaction (HCI) featured highly in the related industries.  ...."   (Considerably more at the link)

Friday, September 28, 2018

Open source Monitoring with Prometheus

Open source monitoring,  good detail at the link.   The word monitoring by itself always attracts me, it means we can understand things in operation, and better understand how to  improve that.  First I had heard of this example.

5 examples of Prometheus monitoring success

Organizations are benefiting from the open source monitoring toolkit's customization, simplicity, and cost savings.

Prometheus is an open source monitoring and alerting toolkit for containers and microservices. The project is a hit with lots of different organizations regardless of their size or industrial sector. The toolkit is highly customizable and designed to deliver rich metrics without creating a drag on system performance. Based on the organizations that have adopted it, Prometheus has become the mainstream, open source monitoring tool of choice for those that lean heavily on containers and microservices.

Conceived at SoundCloud in 2012, Prometheus became part of the Cloud Native Computing Foundation (CNCF) in 2016 and in August 2018, CNCF announced Prometheus was the second "graduated" project in the organization's history.

Prometheus provides a key component for a modern DevOps workflow: keeping watch over cloud-native applications and infrastructure, including another popular CNCF project, Kubernetes. ...."

Thursday, August 30, 2018

Open Source Reinforcement Training

More open source resources here useful for training.    The frameworks involved here are likely the most instructive part.  And for tests to determine comparisons of speed and accuracy.  Note also the criticism of reinforcement learning, while conceptually a very powerful idea,  can be hard to calibrate for general problems.

Google releases open source reinforcement learning framework for training AI models
 Kyle Wiggers   @KYLE_L_WIGGERS  in Venture Beat

Reinforcement learning — an artificial intelligence (AI) technique that uses rewards (or punishments) to drive agents in the direction of specific goals — trained the systems that defeated Alpha Go world champions and mastered Valve’s Dota 2. And it’s a core part of Google subsidiary DeepMind’s deep Q-network (DQN), which can distribute learning across multiple workers in the pursuit of, for example, achieving “superhuman” performance in Atari 2600 games. The trouble is, reinforcement learning frameworks take time to master a goal, tend to be inflexible, and aren’t always stable.

That’s why Google is proposing an alternative: an open source reinforcement framework based on TensorFlow, its machine learning library. It’s available from Github starting today.  .... "