Skip to main content
Python auto-instrumentation library for LlamaIndex. These traces are fully OpenTelemetry compatible and can be sent to an OpenTelemetry collector for viewing, such as Arize Phoenix.

Installation

pip install openinference-instrumentation-llama-index

Compatibility

llama-index versionopeninference-instrumentation-llama-index version
0.12.3+4.0+
0.11.0+3.0+
0.10.43+2.0 - 3.0
0.10.0 - 0.10.431.0 - 0.2
0.9.14 - 0.10.00.1.3

Quickstart

Install packages needed for this demonstration:
python -m pip install --upgrade \
    openinference-instrumentation-llama-index \
    opentelemetry-sdk \
    opentelemetry-exporter-otlp \
    "opentelemetry-proto>=1.12.0" \
    arize-phoenix

Start Phoenix server

Start the Phoenix app in the background as a collector. By default, it listens on http://localhost:6006:
python -m phoenix.server.main serve
The Phoenix app does not send data over the internet. It only operates locally on your machine.

Setup instrumentation

from openinference.instrumentation.llama_index import LlamaIndexInstrumentor
from opentelemetry.exporter.otlp.proto.http.trace_exporter import OTLPSpanExporter
from opentelemetry.sdk import trace as trace_sdk
from opentelemetry.sdk.trace.export import SimpleSpanProcessor

endpoint = "http://127.0.0.1:6006/v1/traces"
tracer_provider = trace_sdk.TracerProvider()
tracer_provider.add_span_processor(SimpleSpanProcessor(OTLPSpanExporter(endpoint)))

LlamaIndexInstrumentor().instrument(tracer_provider=tracer_provider)

Download a text file

import tempfile
from urllib.request import urlretrieve
from llama_index.core import SimpleDirectoryReader

url = "https://raw.githubusercontent.com/Arize-ai/phoenix-assets/main/data/paul_graham/paul_graham_essay.txt"
with tempfile.NamedTemporaryFile() as tf:
    urlretrieve(url, tf.name)
    documents = SimpleDirectoryReader(input_files=[tf.name]).load_data()

Configure OpenAI credentials

import os

os.environ["OPENAI_API_KEY"] = "<your openai key>"

Query the indexed documents

from llama_index.core import VectorStoreIndex

query_engine = VectorStoreIndex.from_documents(documents).as_query_engine()
print(query_engine.query("What did the author do growing up?"))
Visit the Phoenix app at http://localhost:6006 to see the traces.

More Info

Build docs developers (and LLMs) love