Project Configuration

.env File

This file contains the necessary API keys required for the application to function properly. Obtain the API keys from the following sources:

This JSON file holds crucial configuration values for the entire application. Please refer to the documentation before modifying any configurations.

PINECONE_INDEX_NAME: The name of the index, the highest-level organizational unit of vector data in Pinecone.
PINECONE_VECTOR_DIMENSION: Dimensionality of the embedding model's vectors.
PINECONE_UPSERT_BATCH_LIMIT: Number of transcript rows inserted into Pinecone Serverless in parallel.
PINECONE_TOP_K_RESULTS: Number of results fetched by Pinecone for a query.
PINECONE_DELTA_WINDOW: Conversation window size fetched for TOP_K results.
PINECONE_CLOUD_PROVIDER: Cloud provider for Pinecone DB.
PINECONE_REGION: Region of the Pinecone Cloud provider.
PINECONE_METRIC: Distance metric used by Pinecone to calculate similarity.
PINECONE_NAMESPACE: Logical separation inside the Pinecone Index.

EMBEDDING_PROVIDER: Provider of the embedding model for text-to-vector conversion.
EMBEDDING_MODEL_NAME: Name of the embedding model provided by the provider.

LC_LLM_TEMPERATURE: Temperature value for the Large Language Model.
LC_CONV_BUFFER_MEMORY_WINDOW: Conversation memory window limit. (Future Use)
LC_LLM_SUMMARY_MAX_TOKEN_LIMIT: Maximum tokens allowed for summary in the memory buffer.
LC_LLM_MODEL: Large Language Model used for inference.

The config/requirements.txt file was generated using pip freeze, if the primary ./requirements.txt fails, please use config/requirements.txt