Monitoring LangChain

Instrumenting LangChain Application

To instrument the LangChain application, complete the following steps:

Make sure that your environment meets all the prerequisites. For more information, see Prerequisites.

1. To install dependencies for LangChain, run the following command:

   pip3 install langchain-core==0.3.34 langchain-community==0.3.17
   

2. Export the following credentials to access the watsonX models used in the LangChain sample application:

   export WATSONX_URL="<watsonx-url>"
   export WATSONX_PROJECT_ID="<watsonx-project-id>"
   export WATSONX_API_KEY="<watsonx-api-key>"
   

3. Generate a LangChain sample application by running the following code:

    from langchain_ibm import WatsonxLLM
    from langchain_core.prompts import PromptTemplate
    from traceloop.sdk import Traceloop
    from traceloop.sdk.decorators import task, workflow
    import os
   
    # Initialize Traceloop
    Traceloop.init(app_name="langchain_watsonx_service")
   
    @task(name="initialize_watsonx_model")
    def initialize_model():
    return WatsonxLLM(
            model_id="ibm/granite-13b-instruct-v2",
            url=os.getenv("WATSONX_URL"),
            apikey=os.getenv("WATSONX_API_KEY"),
            project_id=os.getenv("WATSONX_PROJECT_ID"),
            params={
                "decoding_method": "sample",
                "max_new_tokens": 512,
                "min_new_tokens": 1,
                "temperature": 0.5,
                "top_k": 50,
                "top_p": 1,
            }
        )
   
    @task(name="create_prompt_template")
    def create_prompt():
        return PromptTemplate(
            input_variables=["input_text"],
            template="You are a helpful AI assistant. Respond to the following: {input_text}"
        )
   
    @task(name="process_llm_query")
    def process_query(prompt, llm, input_text):
        chain = prompt | llm
        return chain.invoke({"input_text": input_text})
   
    @workflow(name="watsonx_conversation_workflow")
    def run_conversation():
        watsonx_llm = initialize_model()
        prompt_template = create_prompt()
   
        input_text = "Explain the concept of quantum computing in simple terms."
        response = process_query(prompt_template, watsonx_llm, input_text)
        print("Response:", response)
   
    if __name__ == "__main__":
        run_conversation()
   
   
   

4. Run the LangChain application by running the following command:

   python3 ./<langchain-sample-application>.py
   

The LangChain sample app generates a response as shown in the following example:

Response: Quantum computing is a type of computing that uses quantum-mechanical phenomena such as superposition and quantum entanglement to perform calculations. It is different from classical computing, which uses binary digits (0s and 1s) to represent information. In quantum computing, information is encoded in quantum bits, or "qubits," which can exist in a superposition of both 0 and 1 at the same time. This allows quantum computers to perform certain types of calculations much faster than classical computers.

After you configure LangChain monitoring, Instana collects the following traces and metrics from the LangChain application:

To view the metrics collected from LLM that is used in LangChain, see View metrics.

Troubleshooting

You might encounter the following issue while monitoring LangChain based applications: