This guide outlines the steps to set up your environment, install necessary libraries, and structure your project for fine-tuning and inference using Azure ML Service as the backend and ONNX Runtime for samples.
Create and activate a new Conda environment with Python 3.10.12.
conda create -n slmopsenv python==3.10.12
conda activate slmopsenv
Install the required Python libraries from the requirements.txt file.
pip install -r requirements.txt
Your project structure is in the following directories:
|--📁 QA_E2E
|-📁 datasets
|-📁 fine-tuning
|-📁 inferences
|-📁 models-cache
📁 datasets - Store the data that needs fine-tuning as JSON format files.
📁 fine-tuning - : Store Microsoft Olive settings in olive-config.json and save a cache of related steps.
📁 inferences - Store inference models and test results.
📁 models-cache - Save fine-tuned Microsoft Phi-3 mini models