> For the complete documentation index, see [llms.txt](https://docs.zingg.ai/zingg/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.zingg.ai/zingg/versions/enterprise-v0.7.0/stepbystep/createtrainingdata/label.md).

# Labeling Records

Providing user feedback on the training pairs

The label phase opens an *interactive* learner where the user can mark the pairs found by **findTrainingData** phase as matches or non-matches. The **findTrainingData** phase generates edge cases for labeling and the label phase helps the user mark them.

`./zingg.sh --phase label --conf config.json <optional --showConcise=true|false>`

Proceed running **findTrainingData** followed by label phases till you have at least 30-40 positives, and when you see the predictions by Zingg converging with the output you want. At each stage, the user will get different variations of attributes across the records. Zingg performs pretty well with even a small number of training, as the samples to be labeled are chosen by the algorithm itself.

**How much training is enough**

The labelled data is the real configuration for Zingg, through which the different Zingg models get created. Heuristically, 30-40 matching records are a good start before you train and match. Actual numbers vary on the size of data, number of attributes, cluster sizes and performance needs. More training, like going to 500-1000 matches, will surely help with performance on almost any dataset. Point to remember here is - Is the effort worth it :smile:

![Shows records and asks user to mark yes, no, can't say on the cli.](https://3168295817-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FktkJPBR9UvYX0Ql7h8sv%2Fuploads%2Fgit-blob-cf3181376aa6b90efb764bbe1f4340e2c210edf4%2Flabelvertical.gif?alt=media)

The **showConcise** flag when passed to the Zingg command line only shows fields which are **NOT DONT\_USE**


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation by asking a question.

Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters:

```
GET https://docs.zingg.ai/zingg/versions/enterprise-v0.7.0/stepbystep/createtrainingdata/label.md?ask=<question>&goal=<user_goal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.