Open Source foundation models in Azure Machine Learning & optimization techniques behind the scenes

From the creator

This episode highlights the new feature we announced at Microsoft Build around foundation models in Azure Machine Learning, allowing ML professionals to discover, fine-tune, and deploy any open source foundation models from Hugging Face. We'll also dig into model optimization techniques that are packaged in Azure Container for PyTorch including DeepSpeed and ONNX Runtime. Visit the AzureML model catalog to browse and discover open source models - https://ml.azure.com/model/catalog Access sample notebooks for using (Finetune / evaluate / deploy) these models - https://github.com/Azure/azureml-examples/tree/main/sdk/python/foundation-models/system 📌 Let's connect: Parinita Rahi | https://www.linkedin.com/in/parinitaparinita/ Swati Gharse | https://www.linkedin.com/in/swati-gharse-3781a97/ Subscribe to the Open at Microsoft: https://aka.ms/OpenAtMicrosoft Open at Microsoft Playlist: https://aka.ms/OpenAtMicrosoftPlaylist New episode every Tuesday! 00:00:00 Introduction 00:02:19 New model Llama 2 00:08:40 Open-source communities

Published 2023-09-12 · Watch on YouTube ↗
Availability and playback restrictions are controlled by the original channel.