Build applications that make decisions (chatbots, agents, simulations, etc...). Monitor, trace, persist, and execute on your own infrastructure.
Python aiburrchatbot-frameworkdagsgenerative-aigraphshacktoberfestllmopsllmsmlopspersistent-data-structurestate-machinestate-managementvisibility
Amazon Bedrock integration
This example shows how to use Burr’s Bedrock helpers:
BedrockAction— single-step Converse call.BedrockStreamingAction— streaming ConverseStream withApplication.stream_result.
Setup
Install dependencies (from the repo root):
pip install -r examples/integrations/bedrock/requirements.txtConfigure AWS credentials and a region where Bedrock is available (for example
AWS_ACCESS_KEY_ID,AWS_SECRET_ACCESS_KEY, andAWS_DEFAULT_REGION).Ensure your account can invoke the model you choose. Override the default model with
BEDROCK_MODEL_IDif needed (default in code is Claude 3 Haiku on Bedrock).
Run
python examples/integrations/bedrock/application.py
The script runs a non-streaming call, then a streaming call, using two small Burr graphs defined in application() and streaming_application().