Bulk Classification from User-Provided Tags.¶

This tutorial shows how to do classification from user provided tags. This is valuable when you want to provide services that allow users to do some kind of classification.

Motivation

Imagine allowing the user to upload documents as part of a RAG application. Oftentimes, we might want to allow the user to specify an existing set of tags, give descriptions, and do the classification for them.

Defining the Structures¶

One of the easy things to do is to allow users to define a set of tags in some kind of schema and save that in a database. Here's an example of a schema that we might use:

tag_id	name	instructions
0	personal	Personal information
1	phone	Phone number
2	email	Email address
3	address	Address
4	Other	Other information

tag_id — The unique identifier for the tag.
name — The name of the tag.
instructions — A description of the tag, which can be used as a prompt to describe the tag.

Implementing the Classification¶

In order to do this we'll do a couple of things:

We'll use the instructor library to patch the openai library to use the AsyncOpenAI client.
Implement a Tag model that will be used to validate the tags from the context. (This will allow us to avoid hallucinating tags that are not in the context.)
Helper models for the request and response.
An async function to do the classification.
A main function to run the classification using the asyncio.gather function to run the classification in parallel.

If you want to learn more about how to do bad computations, check out our post on AsyncIO here.

import openai
import instructor

client = instructor.patch(
    openai.AsyncOpenAI(),
)

First, we'll need to import all of our Pydantic and instructor code and use the AsyncOpenAI client. Then, we'll define the tag model along with the tag instructions to provide input and output.

This is very helpful because once we use something like FastAPI to create endpoints, the Pydantic functions will serve as multiple tools:

A description for the developer
Type hints for the IDE
OpenAPI documentation for the FastAPI endpoint
Schema and Response Model for the language model.

from typing import List
from pydantic import BaseModel, ValidationInfo, model_validator

class Tag(BaseModel):
    id: int
    name: str

    @model_validator(mode="after")
    def validate_ids(self, info: ValidationInfo):
        context = info.context
        if context:
            tags: List[Tag] = context.get("tags")
            assert self.id in {
                tag.id for tag in tags
            }, f"Tag ID {self.id} not found in context"
            assert self.name in {
                tag.name for tag in tags
            }, f"Tag name {self.name} not found in context"
        return self


class TagWithInstructions(Tag):
    instructions: str


class TagRequest(BaseModel):
    texts: List[str]
    tags: List[TagWithInstructions]


class TagResponse(BaseModel):
    texts: List[str]
    predictions: List[Tag]

Let's delve deeper into what the validate_ids function does. Notice that its purpose is to extract tags from the context and ensure that each ID and name exists in the set of tags. This approach helps minimize hallucinations. If we mistakenly identify either the ID or the tag, an error will be thrown, and the instructor will prompt the language model to retry until the correct item is successfully extracted.

@model_validator(mode="after")
def validate_ids(self, info: ValidationInfo):
    context = info.context
    if context:
        tags: List[Tag] = context.get("tags")
        assert self.id in {
            tag.id for tag in tags
        }, f"Tag ID {self.id} not found in context"
        assert self.name in {
            tag.name for tag in tags
        }, f"Tag name {self.name} not found in context"
    return self

Now, let's implement the function to do the classification. This function will take a single text and a list of tags and return the predicted tag.

async def tag_single_request(text: str, tags: List[Tag]) -> Tag:
    allowed_tags = [(tag.id, tag.name) for tag in tags]
    allowed_tags_str = ", ".join([f"`{tag}`" for tag in allowed_tags])

    return await client.chat.completions.create(
        model="gpt-4-turbo-preview",
        messages=[
            {
                "role": "system",
                "content": "You are a world-class text tagging system.",
            },
            {"role": "user", "content": f"Describe the following text: `{text}`"},
            {
                "role": "user",
                "content": f"Here are the allowed tags: {allowed_tags_str}",
            },
        ],
        response_model=Tag,  # Minimizes the hallucination of tags that are not in the allowed tags.
        validation_context={"tags": tags},
    )


async def tag_request(request: TagRequest) -> TagResponse:
    predictions = await asyncio.gather(
        *[tag_single_request(text, request.tags) for text in request.texts]
    )
    return TagResponse(
        texts=request.texts,
        predictions=predictions,
    )

Notice that we first define a single async function that makes a prediction of a tag, and we pass it into the validation context in order to minimize hallucinations.

Finally, we'll implement the main function to run the classification using the asyncio.gather function to run the classification in parallel.

tags = [
    TagWithInstructions(id=0, name="personal", instructions="Personal information"),
    TagWithInstructions(id=1, name="phone", instructions="Phone number"),
    TagWithInstructions(id=2, name="email", instructions="Email address"),
    TagWithInstructions(id=3, name="address", instructions="Address"),
    TagWithInstructions(id=4, name="Other", instructions="Other information"),
]

# Texts will be a range of different questions.
# Such as "How much does it cost?", "What is your privacy policy?", etc.
texts = [
    "What is your phone number?",
    "What is your email address?",
    "What is your address?",
    "What is your privacy policy?",
]

# The request will contain the texts and the tags.
request = TagRequest(texts=texts, tags=tags)

# The response will contain the texts, the predicted tags, and the confidence.
response = asyncio.run(tag_request(request))
print(response.model_dump_json(indent=2))

Which would result in:

{
  "texts": [
    "What is your phone number?",
    "What is your email address?",
    "What is your address?",
    "What is your privacy policy?"
  ],
  "predictions": [
    {
      "id": 1,
      "name": "phone"
    },
    {
      "id": 2,
      "name": "email"
    },
    {
      "id": 3,
      "name": "address"
    },
    {
      "id": 4,
      "name": "Other"
    }
  ]
}

What happens in production?¶

If we were to use this in production, we might expect to have some kind of fast API endpoint.

from fastapi import FastAPI

app = FastAPI()


@app.post("/tag", response_model=TagResponse)
async def tag(request: TagRequest) -> TagResponse:
    return await tag_request(request)

Since everything is already annotated with Pydantic, this code is very simple to write!

Where do tags come from?

I just want to call out that here you can also imagine the tag spec IDs and names and instructions for example could come from a database or somewhere else. I'll leave this as an exercise to the reader, but I hope this gives us a clear understanding of how we can do something like user-defined classification.

Improving the Model¶

There's a couple things we could do to make this system a little bit more robust.

Use confidence score:

class TagWithConfidence(Tag):
    confidence: float = Field(
        ...,
        ge=0,
        le=1,
        description="The confidence of the prediction, 0 is low, 1 is high",
    )

Use multiclass classification:

Notice in the example we use Iterable[Tag] vs Tag. This is because we might want to use a multiclass classification model that returns multiple tag!

await client.chat.completions.create(
    model="gpt-4-turbo-preview",
    messages=[
        {
            "role": "system",
            "content": "You are a world-class text tagging system.",
        },
        {"role": "user", "content": f"Describe the following text: `{text}`"},
        {
            "role": "user",
            "content": f"Here are the allowed tags: {allowed_tags_str}",
        },
    ],
    response_model=Iterable[Tag],
    validation_context={"tags": tags},
)