MLOps Club is building a micro-degree to take you from "Coder" → "Software Engineer" → "Cloud/DevOps Engineer" → "MLOps Engineer". Students have said that our courses feel like going through an rigorous internship with strong mentorship. MLOps is a superset of DevOps and heavily overlaps with Cloud Engineering and Data Engineering. Conceptually, there is a lot to learn to practice MLOps. And the MLOps tooling landscape is crowded and confusing. Many learning resources come from vendors. So, they commonly ignore their own weaknesses, and recommend their own tools even when they are not the best fit. MLOps Club emphasizes "principles first". So that you can build ML Platforms on how to build with popular OSS, Open Core, and closed-source tools. In order of priority, MLOps Club optimizes teaching for 1. covering unchanging principles (the "why") behind the tools/practices recommended 2. building for production and after 3. making DevEx and UX delightful 4. minimizing maintenance and technology cost 5. minimizing tool/vendor lock in Whether you are a learner or creator, welcome to the club!
MLOps Club的外部链接
MLOps Club转发了
Happy Diwali!!! Here's a code for 80% off my "Taking Python to Production" course. It's been a bestseller and "highest rated" course since the day it was launched. If you were joining a high-performing team doing data engineering, data science, or analytics engineering, this would be the onboarding guide. It covers the fundamental software-development skills needed to collaborate and think like an engineer.
MLOps Club转发了
Senior ML/AI Engineer ? MLOps ? Founder @ Decoding ML ~ Posts and articles about building production-grade ML/AI systems.
The truth about being a top ML & MLOPs engineer is not sexy. In reality, you need 2 skills that have nothing to do with AI: 1. ??????????????????????: Learning Python is a perfect start. 2. ?????????? ??????????????????????: Learning to deploy your apps to a cloud such as AWS is critical in bringing value to your project. ?????? ???????? ?????????? 2 ???????????? ?????????????? ???? ???????????? ???? ????????????????????. That's where the ?????????? ?????????????????????? ?????? ???????????? ???????????????????? self-paced course made by ?? Eric Riddoch kicks in. I have known Eric for almost a year, and I can certify that he is a brilliant cloud, DevOps and MLOps engineer. . ???????? ???? ???????? ?????? ???????? ?????????? ???????????? ?????? ????????????: - enterprise-level AWS account management - the fundamentals of cloud engineering - design cloud-native RESTful APIs that cost-effectively scale from 4 to 4 million requests/day - write, test, locally mock, and deploy code using the AWS SDK and OpenAI - advanced observability and monitoring: logs, metrics, traces, and alerts The principles learned during the course can easily be extrapolated from AWS to other cloud platforms such as GCP and Azure. The self-paced course provides Q&A support, a community, and a final live project demo to boost accountability. I strongly recommend it if you want to learn cloud engineering. . To ?????? ????% ??????, use the following ???????? when registering: DECODINGML Eric also offers a ?????????????????????? ?????????????? that can ?????????????????????????? ???????????? ?????? ??????????. →?? ???????????? ????????: https://lnkd.in/dC_dJfiy If you consider buying the course, please use the DECODINGML code to support my work. #machinelearning #mlops #artificialintelligence . ?? Follow me for daily content on production ML and MLOps engineering.
This is the first course in a series called the “MLOps Microdegree”. The teaching style is unique. Both broad and deep, focused on principles, workflows, e2e projects, and career advice.
"Taking Python to Production" just crossed 3k students! ?? ?? (discount code below :) For a lot of new Python devs--especially data scientists and data analysts--software engineering skills are a constant blocker. This course will make them into a strength, rather than a weakness. It covers not just the soup of tools, but also how to bring them together into a super productive workflow (git, GitHub, VS Code, autocompletion, "shipping" code to prod, CI/CD, testing, packaging, versioning, and much more) Udemy is listing it for $75 USD Here's a code to get an 80% discount: PY2PROD3000 Link: https://lnkd.in/gaPT3nBB PS: if you know about the current ruff and uv craze (they're amazing) this course will give you the background to get the absolute most out of them :D
Astral's uv project is finally unifying the fragmented Python ecosystem and it's really good! Some features currently in preview: - platform-agnostic lockfiles (PDM and poetry had this) - `uv publish` (replaces twine) It's fast. It gets a zillion commits/day. Some of the early problems are already gone. No dependencies (unlike Poetry with 60+). It's hooked me on: - using lockfiles - using a .python-version file - rapidly installing dependencies thanks to it's pnpm-like cache Is it ready? Strong "yes" for local development. You can always fall back to pip if a certain thing won't install and use it to manage *everything* else. Medium "yes" for CI/CD. It works most of the time, e.g. when building Docker images. But if you want 100% success in CI/CD, then use uv for development, and pip for prod.
MLOps Club转发了
Amit and I have spent the last 3 months working full time on this monitoring section. Here's one of the first videos! Genuinely, I think this is will be the single best resource online for Python developers to learn to be great at observability without getting turned around or oversold by an expensive vendor... like me. We'll cover 1. (structured) logging 2. traces (autoinstrumentation, correlation, profiling) 3. metrics (log formats, queries, structured) 4. querying alerts and dashboards 5. how to not go bankrupt on log/metric/trace ingestion 6. introducing OpenTelemetry (FREEDOM!!!) and why it's actually awful on AWS 7. generally how to reason about monitoring any system Every time I'd try to "get good at monitoring", I'd run out of time to learn how to do it well, so I was never happy with the result, and what I set up often wasn't used. Will likely be between 50 and 80 total lectures on this. Will be totally hands-on, but also as deep as we can go on the theory. This will be so great!
We just launched a self-paced version of our Python/AWS bootcamp ?? It's comprehensive and extremely up to date. ?? Weekly office hours on Zoom with Eric (Wednesdays at 8am MDT) ?? Cut the price by 20% to reflect the drop from 2 live sessions/week to 1. ?? TA, instructor, and classmate help in Discord ?? Scholarships, discounts, and referral codes available. ?? 25 hours of video, dozens of hands-on labs ?? Refunds available. ?? Overwhelmingly positive reviews. Zero refunds requested so far. Note 1: Self-paced is in experimental beta. We're pausing leading live cohorts through December and will reevaluate pricing and delivery formats in 2025. Note 2: The principle behind financial aid. We want to ?? accessibly priced for students and ?? those who have been laid off or are struggling to find a job. ?? We also want to encourage word-of-mouth referrals. Hence, the generous referral program. ?? encourage folks to build a SWE foundation and make it more likely they'll succeed, hence the $100 discount for completing our highest-rated "Taking Python to Production" on Udemy first The "Course Overview" video is a bit out of date (we've added more), so we'll be updating that in the next few weeks. More details on mlops-club(dot)org cc: Amit Vikram Raj, Mert Bozkir, Maria Vechtomova, Paul Iusztin, Meri Nova
Eric Riddoch is giving a talk at the Data Engineering and Machine Learning Summit 2024 (organized by Benjamin Rogojan and Xinran Waibel) this ???????????? ???? 13.00 ??????! ?? He'll be sharing insights from over 4 years of experience building ML platforms and addressing key questions like: - What exactly is an ML platform? - How does platform engineering fit into MLOps. - Why the "handoff model" between DS and DE teams often fails. - Key principles for building a strong platform. ?? Register for the conference and session here: https://lnkd.in/ebD6_W3A Plus, AI Legend Abi Aryan will be speaking about Data Management for LLMs!
Linkedin's search feature is //awful// Let's index our Linkedin posts using both vector embeddings and structured data (comments, reactions, mentions) and make them searchable on AWS for FREE! Maybe a chatbot will follow. Sentence transformers and sqlite-vec look like they'll be great for this :D
此处无法显示此内容
在领英 APP 中访问此内容等
First live stream is on the books! ?? My idea here is: I always find it helpful to watch experienced engineers' workflow. I've led a world-class ML Platform team at the staff level. Maybe there's something in my workflow or tool choices that's useful for you :) ... or maybe you'll see this is pretty standard stuff--not as unreachable as you thought. In this stream, I: ? created a sandbox AWS account just for this project ? bootstrapped an AWS CDK project ? deployed a Python lambda to AWS with some advanced dependency management Next Tuesday, I'll index my Linkedin posts in a vector database and used today's work to serve them... for FREE :D If you want to join some of these: ?? join live and fire questions on ANY TOPIC in the chat ?? or watch the edited version later (about a third shorter)