Pipeline class - Constructor

TypeScript

Python

The Pipeline class is the main entry point for passing observability data to Gentrace. This document focuses on the constructor. Learn more about using pipelines [here] (pipeline/basics).

Constructor

Arguments

options

The constructor takes an options object with the type PipelineOptions.

Types

🛠️ PipelineOptions

typescript

slug?

An optional slug for the pipeline.

logger?

An optional logger object with info and warn methods for logging messages and warnings.

typescript
import { Pipeline} from "@gentrace/core";
 
const pipeline = new Pipeline({
  slug: "my-pipeline",
  logger: {
    info: (message, context) => {
      console.log(message, context);
    },
    warn: (message, context) => {
      console.warn(message, context);
    }
  }  
});
 

plugins?

An optional object containing the plugins to be used within this pipeline.

typescript
import { Pipeline } from "@gentrace/core";
import { initPlugin } from "@gentrace/openai";
 
const openaiPlugin = await initPlugin({
  apiKey: process.env.OPENAI_KEY,
});
 
const pipeline = new Pipeline({
  slug: "my-pipeline",
  logger: {
    info: (message, context) => {
      console.log(message, context);
    },
    warn: (message, context) => {
      console.warn(message, context);
    }
  },
  plugins: {
    openai: openaiPlugin
  }
});

python

slug?

An optional slug for the pipeline.

openai_config?

The OpenAI configuration object, so the Pipeline can properly instantiate an OpenAI handle instance.

python
import os
import gentrace
gentrace.init(
    api_key=os.getenv("GENTRACE_API_KEY"),
)
pipeline = gentrace.Pipeline(
    "example-pipeline",
    # ✏️ Add your OpenAI configuration here
    openai_config={
        "api_key": os.getenv("OPENAI_KEY"),
    },
)
pipeline.setup()
runner = pipeline.start()
# ✅ Defining the openai_config parameter allows you to access the OpenAI handle instance
openai = runner.get_openai()

pinecone_config?

The Pinecone configuration object, so the Pipeline can properly instantiate a Pinecone handle instance.

python
import os
import gentrace
gentrace.init(
    api_key=os.getenv("GENTRACE_API_KEY"),
)
pipeline = gentrace.Pipeline(
    "example-pipeline",
    # ✏️ Add your OpenAI configuration here
    pinecone_config={
        "api_key": os.getenv("PINECONE_API_KEY"),
        "environment: os.getenv("PINECONE_ENVIRONMENT"),
    },
)
pipeline.setup()
runner = pipeline.start()
# ✅ Defining the openai_config parameter allows you to access the OpenAI handle instance
openai = runner.get_openai()