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
fixed typos and broken link
ziltod2b67f90d0e8
diff --git a/examples/rag-lancedb-ingestion/README.md b/examples/rag-lancedb-ingestion/README.mdindex 312e4d43..65098ef2 100644--- a/examples/rag-lancedb-ingestion/README.md+++ b/examples/rag-lancedb-ingestion/README.md@@ -6,9 +6,9 @@ This example shows how to build a chatbot with RAG over Substack blogs (or any R> Burr UI brings a new level of observability to your RAG application via OpenTelemetry-Burr + [LanceDB](https://lancedb.github.io/lancedb/) constitute a powerful, but lightweight combo to build retrieval-augmented generative (RAG) agents. Burr allows to define complex agents in an easy-to-understand and debug manner. It also provides all the right features to help you productionize agents including: monitoring, storing interactions, streaming, and a powerful open-source UI.+Burr + [LanceDB](https://lancedb.github.io/lancedb/) constitute a powerful, but lightweight combo to build retrieval-augmented generative (RAG) agents. Burr allows to define complex agents in an easy-to-understand and debug manner. It also provides all the right features to help you productionize agents including: monitoring, storing interactions, streaming, and a fully-featured open-source UI.-LanceDB makes it easy to swap embedding provider, and hides this concerns from the Burr application layer. For this example, we'll be using [OpenAI](https://github.com/openai/openai-python) for embedding and response generation.+LanceDB makes it easy to swap embedding providers, and hides this concern from the Burr application layer. For this example, we'll be using [OpenAI](https://github.com/openai/openai-python) for embedding and response generation.By leveraging the [Burr integration with OpenTelemetry](https://blog.dagworks.io/p/building-generative-ai-agent-based), we get full visibility into the OpenAI API requests/responses and the LanceDB operations for free.diff --git a/examples/rag-lancedb-ingestion/notebook.ipynb b/examples/rag-lancedb-ingestion/notebook.ipynbindex 8032eed7..6a070772 100644--- a/examples/rag-lancedb-ingestion/notebook.ipynb+++ b/examples/rag-lancedb-ingestion/notebook.ipynb@@ -113,7 +113,7 @@"2. Chunk the text string by splitting it into sentences\n","3. Join sentence chunks into larger \"context windows\" via a rolling operation.\n","\n",- "We use a custom trick to map and store the relationship between HTML pages, sentence chunks, and context windows ([learn more](\"https://github.com/dlt-hub/dlt/issues/1699\"))."+ "We use a custom trick to map and store the relationship between HTML pages, sentence chunks, and context windows ([learn more](https://github.com/dlt-hub/dlt/issues/1699))."]},{