Python · Instrumentation
openinference-instrumentation-llama-index
OpenInference LlamaIndex Instrumentation
- Module
openinference.instrumentation.llama_index- Version
- 4.7.7
- Python
- >=3.10, <3.16
- Instruments
llama-index-core >= 0.14.19, llama-index-multi-modal-llms-openai >= 0.1.7
Install
pip install openinference-instrumentation-llama-indexUsage
Instrument once, at startup, before the library is used. The LlamaIndexInstrumentor reference below lists the options.
from opentelemetry.sdk.trace import TracerProvider
from openinference.instrumentation import TraceConfig
from openinference.instrumentation.llama_index import LlamaIndexInstrumentor
tracer_provider = TracerProvider()
# ...add a span processor that exports to Phoenix or any OTLP backend.
LlamaIndexInstrumentor().instrument(
tracer_provider=tracer_provider,
config=TraceConfig(hide_inputs=True), # optional
)openinference.instrumentation.llama_index
LlamaIndexInstrumentor
classSource ↗class LlamaIndexInstrumentor(BaseInstrumentor)An instrumentor for LlamaIndex
instrument
methoddef instrument(
*,
tracer_provider: Optional[TracerProvider] = None,
config: Optional[TraceConfig] = None,
skip_dep_check: bool = False,
) -> NoneStart tracing llama-index-core, llama-index-multi-modal-llms-openai: patch it so its calls make OpenInference spans. Calling it again does nothing. Inherited from OpenTelemetry’s BaseInstrumentor.
Parameters
| Name | Description |
|---|---|
tracer_provider | The tracer provider the spans are made with. Defaults to the global one, |
config | What the spans record: hide inputs, outputs or messages, or limit image sizes. Defaults to |
skip_dep_check | Instrument even if the installed version of |
uninstrument
methoddef uninstrument() -> NoneStop tracing: undo the patches instrument() made. Inherited from OpenTelemetry’s BaseInstrumentor.
instrumentation_dependencies
methodSource ↗def instrumentation_dependencies() -> Collection[str]get_current_span
functionSource ↗def get_current_span() -> Optional[Span]