Skip to content

Configuration.md

Latest commit

 

History

History
35 lines (32 loc) · 2.53 KB

Configuration.md

File metadata and controls

35 lines (32 loc) · 2.53 KB
Raw

Configuration

The SDK reads its configuration from environment variables. Every variable is optional — with no configuration the SDK runs against an in-memory store. The most important ones are:

Variable Purpose Default
AGENT_LEARNING_STORE_BACKEND Storage backend: memory, cosmos, or local memory
AGENT_LEARNING_COSMOS_ENDPOINT Cosmos DB account URL (only used when backend is cosmos) unset
AGENT_LEARNING_COSMOS_DATABASE Cosmos DB database name (only used when backend is cosmos) dq_rl
AGENT_LEARNING_LOCAL_STORE_DIR Directory for the local file backend ./data/agent-learning/store
AGENT_LEARNING_SCORE_MODE Compatibility scorer mode: nlp or llm llm
AGENT_LEARNING_SCORE_TIER Preferred scorer tier: stdlib, nlp, slm, or llm unset
AGENT_LEARNING_SCORE_ENDPOINT Azure OpenAI endpoint used by the scorer unset
AGENT_LEARNING_SCORE_DEPLOYMENT Scorer deployment name unset
AGENT_LEARNING_SCORE_API_KEY API key for the LLM scorer unset
AGENT_LEARNING_SCORE_API_VERSION Azure OpenAI API version used by the LLM scorer 2024-10-21
AGENT_LEARNING_SCORE_CREDENTIAL_MODE Azure credential resolution mode for the LLM scorer unset
AGENT_LEARNING_SCORE_USER_ASSIGNED_CLIENT_ID User-assigned managed identity client ID unset
AGENT_LEARNING_SCORE_CREDENTIAL_SCOPE Azure token scope for the LLM scorer https://cognitiveservices.azure.com/.default
AGENT_LEARNING_NLP_SCORE_DIR Snapshot directory for feature-based NLP scorers ./data/agent-learning/nlp-scores
AGENT_LEARNING_STDLIB_SCORE_DIR Snapshot directory for Tier 1 stdlib scorers ./data/agent-learning/stdlib-scores
AGENT_LEARNING_NLP_TEXT_SCORE_DIR Snapshot directory for Tier 2 NLP text scorers ./data/agent-learning/nlp-text-scores
AGENT_LEARNING_W_INTENT Weight for intent-resolution reward 0.4
AGENT_LEARNING_W_ADHERENCE Weight for task-adherence reward 0.3
AGENT_LEARNING_W_COMPLETION Weight for task-completion reward 0.3
AGENT_LEARNING_LR REINFORCE learning rate 0.05
AGENT_LEARNING_BASELINE_DECAY EMA decay on the value baseline 0.9

By default the SDK uses a volatile in-memory store. Set AGENT_LEARNING_STORE_BACKEND=cosmos (together with the Cosmos variables above) for durable Cosmos DB persistence, or =local to persist to JSON files on disk. When the score configuration is missing, the SDK skips evaluations so unit tests still pass.