This sample has the full End2End process of creating RAG application with Prompt Flow and AI Studio. It includes GPT 3.5 Turbo LLM application code, evaluations, deployment automation with AZD CLI, GitHub actions for evaluation and deployment and intent mapping for multiple LLM task mapping.
Promptflow release v1.8.0 (2024.04.10) split the default packages into multiple focused packages, with this guidance for Azure usage: promptflow-azure: Azure extra requires(promptflow[azure]) for promptflow to integrate with Azure.
Azure OpenAI Chat Completion API requires an api-version string. This was set to 2023-03-15-preview but that value will be retired on Jul 1, 2024 per documentation. Fix replaces the hardcoded value with a dynamically determined value that can be set from env variable (and defaults to newer "2024-03-01-preview" if not set)
Changes were validated with deploy to Azure from Codespaces.
Promptflow release v1.8.0 (2024.04.10) split the default packages into multiple focused packages, with this guidance for Azure usage:
promptflow-azure: Azure extra requires(promptflow[azure]) for promptflow to integrate with Azure.
Azure OpenAI Chat Completion API requires an
api-version
string. This was set to2023-03-15-preview
but that value will be retired on Jul 1, 2024 per documentation. Fix replaces the hardcoded value with a dynamically determined value that can be set from env variable (and defaults to newer "2024-03-01-preview" if not set)Changes were validated with deploy to Azure from Codespaces.
cc: @cassiebreviu for review