- Видео 188
- Просмотров 53 282
Comet ML
Добавлен 13 авг 2018
Comet is doing for ML what GitHub did for code. We allow data science teams to automatically track their datasets, code changes, experimentation history and production models creating efficiency, transparency, and reproducibility. Our vision is to allow every company to harness the power of AI.
Learn more & see a demo at www.comet.ml/
Connect with us on social media:
• Join the conversation on our community publication Heartbeat ( www.heartbeat.comet.ml)
• Follow us on Twitter ( cometml )
• Connect with us on LinkedIn ( www.linkedin.com/company/comet-ml )
• Like our Facebook page ( cometdotml/ )
Learn more & see a demo at www.comet.ml/
Connect with us on social media:
• Join the conversation on our community publication Heartbeat ( www.heartbeat.comet.ml)
• Follow us on Twitter ( cometml )
• Connect with us on LinkedIn ( www.linkedin.com/company/comet-ml )
• Like our Facebook page ( cometdotml/ )
Introducing Opik: Open-Source LLM Evaluation from Comet
www.comet.com | The Comet team is excited to introduce Opik, an end-to-end LLM evaluation platform for developers building LLM-powered applications. With Opik’s LLM observability tooling across your development and production lifecycle, you can build, test, ship, and continuously improve your applications with confidence.
In this quick intro video, Head of Product Jacques Verré walks through common LLM app development challenges and how to begin solving them with Opik. It all starts with better visibility into your prompt engineering inputs and results: log and annotate traces in development and understand the outcomes generated by your system prompts and test data. Then, perform unit tes...
In this quick intro video, Head of Product Jacques Verré walks through common LLM app development challenges and how to begin solving them with Opik. It all starts with better visibility into your prompt engineering inputs and results: log and annotate traces in development and understand the outcomes generated by your system prompts and test data. Then, perform unit tes...
Просмотров: 1 422
Видео
Leveraging MLOps to Speed Up Historical Document Processing
Просмотров 265Год назад
Stanley Fujimoto, a senior data scientist at Ancestry, discusses how his team leveraged MLOps and collaboration to quickly process and extract information from 6.6 million historical census images. He explains their multi-model pipeline using deep learning for document layout extraction and handwriting recognition. Key themes include the importance of infrastructure, transparency, and calibrate...
From Hackathon to Production: Unveiling Etsy's Image Search Revolution with Comet
Просмотров 391Год назад
In this engaging fireside chat, Gideon, CEO of Comet, Eden Dolev, Senior ML-II Engineer, and Alaa Awad Staff ML Software Engineer at Etsy, come together to discuss the development and implementation of an innovative image search product at Etsy. Try Comet for free today: www.comet.com/signup?RUclips&Etsy-event Learn more about Comet for Enterprise: www.comet.com/site/enterprise/?RUclips&Etsy-event
LLMOps+ Model CI/CD with Comet
Просмотров 654Год назад
Leverage the latest product enhancements around LLMOps such as Prompt Visualizations and Integrations with OpenAI and LangChain for your upcoming LLM projects. This video also covers other critical aspects of the Comet platform like: - Model CI/CD Workflows: Streamline your update processes for models in production. - Enhanced CV Model Predictions Visualization: Gain deeper insights into your C...
How Netflix Built Their ML Infrastructure
Просмотров 2,2 тыс.Год назад
This fireside chat between Prasanna Padmanabhan, ML Platform Director at Netflix, and Gideon Mendels, CEO at Comet, delves into the journey of building an MLOps team at Netflix with the help of Comet. In this engaging session, they will give you a glimpse into the intricate architecture of Netflix’s machine learning infrastructure and share valuable insights into the intricacies of leading MLOp...
Comet Interactive Confusion Matrix
Просмотров 302Год назад
Easily spot misclassifications, analyze pattern irregularities in your data and identify problematic samples or irregularities in your ML datasets. Interact with a Comet public workspace here: www.comet.com/kristenkehrer/dogs-and-cats/view/new/panels?RUclips& Sign up for an account here: www.comet.com/signup?RUclips&
Comet Custom Panels
Просмотров 223Год назад
Build bespoke visualizations for your experiment data that are tailored to your use case. Use our Python or Javascript Panel SDK’s to build sophisticated, interactive tools that extend Comet’s visualization capabilities to new problems. Interact with a Comet public workspace here: www.comet.com/kristenkehrer/dogs-and-cats/view/new/panels?RUclips& Sign up for an account here: www.comet.com/signu...
Three things you can do with Comet and Gradio
Просмотров 482Год назад
Comet integrates with Gradio! Gradio is an open-source Python library that lets machine learning developers create demos and GUIs from models very easily, with just a few lines of Python. Comet allows data scientists to track their machine learning experiments at every stage, from training to production, and interactively engage with the dashboard. Together, you can collaborate across your team...
You can watch your model train real-time in Comet, or... #shorts
Просмотров 51Год назад
You can watch your model train real-time in Comet, or... #shorts
Scaling ML Operations for a Multi-Sided Retail Marketplace: How Shipt Leverages Comet
Просмотров 211Год назад
Scaling ML Operations for a Multi-Sided Retail Marketplace: How Shipt Leverages Comet
Creating an asthma attack prediction model
Просмотров 188Год назад
Creating an asthma attack prediction model
Convergence Conference: Rochelle March "Building a global data supply chain"
Просмотров 29Год назад
Convergence Conference: Rochelle March "Building a global data supply chain"
Convergence Conference: Abubakar Abid "Building Interactive Machine Learning Demos Fast"
Просмотров 33Год назад
Convergence Conference: Abubakar Abid "Building Interactive Machine Learning Demos Fast"
Convergence Conference: Shivika K. Bisen "Testing ML models for production"
Просмотров 39Год назад
Convergence Conference: Shivika K. Bisen "Testing ML models for production"
Convergence Conference: Resham Sarkar "Deep Dish: How ML brings you closer to pizza"
Просмотров 15Год назад
Convergence Conference: Resham Sarkar "Deep Dish: How ML brings you closer to pizza"
Convergence Conference: Kevin Stumpf "How Feature Stores Enable Operational ML"
Просмотров 64Год назад
Convergence Conference: Kevin Stumpf "How Feature Stores Enable Operational ML"
Convergence Conference: Eduardo Bonet "Using ML to Solve the Right Problems"
Просмотров 18Год назад
Convergence Conference: Eduardo Bonet "Using ML to Solve the Right Problems"
Convergence Conference: Peter Gao "It's The Data, Stupid!"
Просмотров 28Год назад
Convergence Conference: Peter Gao "It's The Data, Stupid!"
Convergence Conference: Panel "How to put ML successfully into production"
Просмотров 97Год назад
Convergence Conference: Panel "How to put ML successfully into production"
Convergence Conference: Uri Goren "Recommendation systems: From A/B testing to deep learning"
Просмотров 43Год назад
Convergence Conference: Uri Goren "Recommendation systems: From A/B testing to deep learning"
Convergence Conference: Vidhi Chugh "Data Quality Assessment using TensorFlow Data Validation"
Просмотров 68Год назад
Convergence Conference: Vidhi Chugh "Data Quality Assessment using TensorFlow Data Validation"
Convergence Conference: Emily Curtin "Stop Making Data Scientists Do Systems"
Просмотров 46Год назад
Convergence Conference: Emily Curtin "Stop Making Data Scientists Do Systems"
How to get investors in regulated markets
Просмотров 142 года назад
How to get investors in regulated markets
can you provide the link of the dataset you are using
Loved the info, thanks for sharing!
If the echo annoys you, just stick with it until around 5:30, when they fix it.
DO you have the code for this?
☝️ "Promosm"
I follow the same instruction, but nothing is monitored automatically. Only parameters that are passed through Command are captures, e.g. epochs. can anyone tell me why this happens?
most of the auto capture is for keras. you can look at their Docs to see what they auto log for your development platform.
毕设留名
Geez, she’s just as boring and dull as before.
best of luck
Great work. I want to use this code Can you please share the github repo or Notebook file for it.
Where is the github Repo for this?
Hey Hammad, here it is: Code: github.com/SkalskiP/sport
Gorgeous! Thanks for having me!
Can you share the link u used it in demo
You can sign up for the Comet platform here: www.comet.com/signup
new viewer, came for Nicola #rstats
This is a very useful platform👏
Nice Kristen!
Slides would have helped a bit
Congratulations! It is nice to know that the Comet platform has become mature now. I had some experiences with years ago, they passed from good to excellent! 👍
If you want to learn more about the book you can visit the book website: www.cometfordatascience.com/
In 2030 we are gonna have flying cars 👌🏼.....Remote first all day all night😁
This is very informative and really nice to watch!
Awesome video 👍 Thank you for sharing your book writing experience. I can't wait to read the book 😀
Just Amazing, Especially those last 7 Minutes. Got to learn a few new things. my question for next week- How different is deploying a model on free cloud and in production?
This is really awesome Thank you so much
What a presentation!!♥️ very helpful !
Amazing content. I love it
Buy when you have money and big needs , build when you have no money and small needs
Really good presentation on Fast.ai callbacks, Sylvain makes it really easy to understand such and important feature of this amazing library!
Great presentation form mister Victor. Hopely more.
Sylvain's talk starts at 7:35
Audio quality really detracts from an otherwise informative presentation.