🪨 feat: AWS Bedrock embeddings support

[FinnConnor]

Added bedrock embeddings support. Need to set env variables 1) AWS_ACCESS_KEY 2) AWS_SECRET_ACCESS_KEY 3) EMBEDDINGS_PROVIDER=bedrock and optionally 3) AWS_DEFAULT_REGION 4)EMBEDDINGS_MODEL (default: amazon.titan-embed-text-v1)

Tested using pgvector vector storage with bedrock embedding provider with model amazon.titan-embed-text-v1 through rag docker image.

1. Successfully uploaded and embedded a small text file
2. Successfully performed query search returning relevant documents.

[FinnConnor]

Tested amazon.titan-embed-text-v2:0 text embedding model through bedrock using pgvector as vector store. Used docker

1. Upload text document and embed into vector storage.
2. Verify able to query relevant documents.

[FinnConnor]

For AWS credentials named env variables:

AWS_DEFAULT_REGION
AWS_ACCESS_KEY_ID
AWS_SECRET_ACCESS_KEY

Two possible changes to consider , adding AWS_SESSION_TOKEN as needed (optional), also prefixing 'BEDROCK' to all env variables.