Skip to main content
Version: 2.0.0

PipelineRun class - Pinecone handler

The PipelineRun class instance exposes a Pinecone handler that simplifies capturing generative output in Gentrace.

For a guided walkthrough of the Gentrace Pinecone integration, visit our quickstart here.

Usage​

When the PipelineRun instance is created, you can create a handle in two different ways to communicate with Pinecone.

Simple​

Creating a simple handle is the easiest way to get started.

typescript
import { init, Pipeline } from "@gentrace/core";
import { Pinecone } from "@gentrace/pinecone";
import { DEFAULT_VECTOR } from "../constants";
 
init({
  apiKey: process.env.GENTRACE_API_KEY
});
 
const pinecone = new Pinecone({
  apiKey: process.env.PINECONE_API_KEY!,
});
 
const index = await pinecone.Index("openai-trec");
 
const queryResponse = await index.query({
  includeValues: true,
  topK: 3,
  vector: DEFAULT_VECTOR,
  pipelineSlug: "pinecone-example",
});
 
console.log("queryResponse", queryResponse);

Advanced​

The more advanced way involves creating a handle that is bound to a PipelineRun instance. This allows you to attach multiple steps to a single PipelineRun instance.

typescript
const runner = pipeline.start();
 
const pinecone = runner.pinecone;
 
const index = await pinecone.Index("openai-trec");
 
const queryResponse = await index.query({
  includeValues: true,
  topK: 3,
  vector: DEFAULT_VECTOR,
});
 
await runner.submit();

Fetch​

The Index.fetch() method is a wrapper around the equivalent Pinecone SDK method.

Arguments​

The original Index.fetch() method parameters are supported. The below key/value pairs overwrite or augment the default key/value pairs for the options object in the first parameter.

pipelineSlug?​

If you're using the Simple SDK, you can specify the pipeline slug here. If you're using the Advanced SDK, this is set in the Pipeline constructor and must be omitted.

gentrace​

This object contains Gentrace context. Learn about context here.

Return value​

The return value is a Promise that resolves to the original OpenAI response. If you're using the simple SDK, the response is augmented with a pipelineRunId that references the run.

pipelineRunId?​

The pipelineRunId is a unique identifier for the run. It can be used to retrieve the run later.

Examples​

typescript
const runner = pipeline.start();
 
const pinecone = runner.pinecone;
 
const index = await pinecone.index("my-index");
 
const response = await index.fetch(["8248"], {
  gentrace: {
    metadata: {
      gitSha: {
        type: "string",
        value: "6cb2a7e4f5765d92d95728f1746b9a1a59d8ad13"
      }
    }
  }
});
 
await runner.submit();

Upsert​

The Index.upsert() method is a wrapper around the equivalent Pinecone SDK method.

Arguments​

The original Index.upsert() method parameters are supported. The below key/value pairs overwrite or augment the defaults.

vectors​

Same as the underlying Pinecone SDK, this parameter can be a single vector or an array of vectors that you want to upsert.

options​

This object contain Gentrace-specific options.

  • pipelineSlug? If you're using the Simple SDK, you can specify the pipeline slug here. If you're using the Advanced SDK, this is set in the Pipeline constructor and must be omitted.

  • gentrace? This object contains Gentrace context. Learn about context here.

Return value​

The return value is a Promise that resolves to the original OpenAI response. If you're using the simple SDK, the response is augmented with a pipelineRunId that references the run.

pipelineRunId?​

The pipelineRunId is a unique identifier for the run. It can be used to retrieve the run later.

Examples​

typescript
const runner = pipeline.start();
 
const pinecone = runner.pinecone;
 
const index = await pinecone.index("my-index");
 
const upsert = await index.upsert([
  {
    id: String(Math.floor(Math.random() * 10000)),
    values: DEFAULT_VECTOR,
  },
]);
 
await runner.submit();

Query​

The Index.query() method is a wrapper around the equivalent Pinecone SDK method.

Arguments​

The original Index.query() method parameters are supported. The below key/value pairs overwrite of augment the defaults for the first parameter.

pipelineSlug?​

If you're using the Simple SDK, you can specify the pipeline slug here. If you're using the Advanced SDK, this is set in the Pipeline constructor and must be omitted.

gentrace?​

This object contains Gentrace context. Learn about context here.

Return value​

The return value is a Promise that resolves to the original OpenAI response. If you're using the simple SDK, the response is augmented with a pipelineRunId that references the run.

pipelineRunId?​

The pipelineRunId is a unique identifier for the run. It can be used to retrieve the run later.

Examples​

typescript
const runner = pipeline.start();
 
const pinecone = runner.pinecone;
 
const index = await pinecone.index("my-index");
 
const queryResponse = await index.query({
  includeValues: true,
  topK: 3,
  vector: DEFAULT_VECTOR,
  pipelineSlug: "pinecone-example",
});
 
await runner.submit();

Delete​

The Index.deleteOne() method is a wrapper around the equivalent Pinecone SDK method.

Arguments​

The original Index.deleteOne() method parameters are supported. The below key/value pairs overwrite or augment the defaults.

vectorId​

Same as the underlying Pinecone SDK, this parameter is the ID of the vector you want to delete.

options​

This object contain Gentrace-specific options.

  • pipelineSlug? If you're using the Simple SDK, you can specify the pipeline slug here. If you're using the Advanced SDK, this is set in the Pipeline constructor and must be omitted.

  • gentrace? This object contains Gentrace context. Learn about context here.

Return value​

The return value is a Promise that resolves to the original OpenAI response. If you're using the simple SDK, the response is augmented with a pipelineRunId that references the run.

pipelineRunId?​

The pipelineRunId is a unique identifier for the run. It can be used to retrieve the run later.

Examples​

typescript
const runner = pipeline.start();
 
const pinecone = runner.pinecone;
 
const index = await pinecone.index("my-index");
 
const response =  await index.deleteOne("8248");
 
await runner.submit();