evaluate the existing label correct or not

I want to evaluate the existing label correct or not. Could you show me what should I do? which command I should use? I’ve already imported them to prodigy.
![existing-label|690x183](uploa

d://kJxctviNv05e2EzS3mKX98OOF3G.png)

Hi @coco,

It looks like the start and end values inside your example spans array are strings instead of numbers because they’re wrapped in " characters.

You have:

{
    ...
    "spans": [
        {
            "start": "42",
            "end": "43"
        }
    ]
}

But I think you want:

{
    ...
    "spans": [
        {
            "start": 42,
            "end": 43
        }
    ]
}

Thank you. I removed the " around the value as you show me. I want to evaluate the existing label correct or not. which command I should use?

What do you mean by “evaluate if the existing label is correct”? Do you want to load in your data and re-annotate it to make sure it’s correct?

I want to load my data with annotation. I have already annotated the text with other app, now I want to import the output from other app to prodigy to validate the output results from other app.

If you just want to look at the entries and the label, you could do this:

python -m prodigy mark test.jsonl --view-id classification

This will show the text and the "label" assigned in the data.

By the way, one note on the examples you posted: It looks like the "text" only contains the text of the entity you labelled and not the full text that the start and end in the "spans" refer to? Is that intended? For example, "start": 188 wouldn’t actually map to a valid entity character offset in the "text", because it’s not actually the full, original text.