Видео

Can we scale LLMs like Llama 3.2 405B models or some other open source models for custom dataset

00:06 Introduction and interactive engagement for ML pipelines workshop. 02:13 Introduction to workflow platform setup and community engagement. 05:49 Introduction to ML Pipelines and community engagement. 07:41 Overview of building and deploying machine learning models in a streamlined workflow. 11:28 Efficient workflows are essential for reliable AI applications. 13:19 Improving efficiency and cost-effectiveness in AI workflows is crucial. 16:57 Optimize resource utilization and ensure type safety in AI workflows. 18:33 Union orchestrates efficient ML workflows using a containerized and scalable approach. 21:51 Guide on using Colab for ML coding with GitHub and Hugging Face. 23:32 Creating an access token on Hugging Face for AI workflows. 27:39 Overview of building a basic ML workflow using tasks and workflows. 29:27 Demonstrating the execution of a basic ML workflow locally and on the cloud. 33:04 Overview of project folders and troubleshooting connection issues. 35:02 Addressing onboarding issues for workshop participants. 38:43 Running a container image build while configuring a new model repository. 40:19 Defining container images simplifies AI workflow management. 43:45 Setting up resource specifications and functions for ML workflows. 45:26 Implementing train-test split for reliable model training and evaluation. 48:56 Uploading model to Hugging Face with API integration. 50:31 Model evaluation can enhance pipeline reliability through additional testing and experimentation. 53:49 Identifying key features greatly enhances model prediction accuracy. 55:37 Batch processing of data using ML models can yield predictions even with out-of-range inputs. 59:14 Demonstrating model downloading and prediction using Hugging Face Hub. 1:01:04 Demonstrating rerunning tasks and accessing execution details using the API. 1:04:53 Utilize Union artifacts for streamlined ML pipeline data management. 1:06:38 Overview of building a minimum ML pipeline for production use. 1:10:52 Managing model versions with Hugging Face improves reliability in ML workflows.

Awesome content man

easy for understand like tire thread gap

How to find distance between 2 object like 2 square on paper find gap between 2 square Please help

Thanks for the video, it would be really helpful to add timestamps to the different topics.

Hello and Thank you Bernease Herman.

Very good content! Keep it up

Thank you for this incredible session. where can I find the github link?

Your content is good but I would suggest to please take care of following if possible: 1. Due to bright window in the background of Kevin his face was not really visible 2. When presenting the code please try to increase the font size to make it easier for us to read the code 3. Relateed to point 2 - please try to upload vidoes in 1080p format

Thank you for this informative session. Very well explained!! For the workflow example that you were running (with version prod-v0) and you were not able to see in console UI, was that because you registered workflow in production domain and i guess you were checking in development ?

Kevin, you are the best.🔥

😒 PЯӨMӨƧM

Thank you. I am humbled 🙏 for the acknowledgment. Hoping to contribute more in the future.