prodigy custom labeling interface

custom_task_template.py

from prodigy.components.filters import filter_duplicates
from prodigy.components.stream import get_stream
from prodigy import recipe
from prodigy import set_hashes

@recipe(
    "custom-text-classification",
    dataset=("The dataset to save to", "positional", None, str),
    source=("Path to the source data as JSONL", "positional", None, str),
)
def custom_text_classification(dataset: str, source: str):
    # Load your source data
    stream = get_stream(source, dedup=True, rehash=True)
    stream = (set_hashes(eg, input_keys=("id",), task_keys=("label",), ignore=("span","text", "input", "html", "image"), overwrite = True) for eg in stream)
    stream = filter_duplicates(stream, by_input=True, by_task=True)
    
    # Define the annotation blocks
    blocks = [
        {
            "view_id": "html",
            "html_template": """
                <div class='custom-container' style="display: flex; font-size: 15px">
                    <div style="flex: 1;">
                        <img src="{{meta.signed_image_url}}" alt="Page Image" style="width: 100%; height: auto;">
                    </div>
                    <div style="flex: 1; padding-left: 10px;">
                        <h3>{{meta.source}}</h3>
                        
                        <p>{{meta.preamble}}</p>
                        <p style="background: #c8eec88a;">{{text}}</p>
                        <p style="font-size;12px; text-align: right;">Page: {{meta.page}}</p>
                    </div>
                </div>
            """
        },
        {
            "view_id": "choice",
            "options": [
                {"id": "Product", "text": "Product"},
                {"id": "Company", "text": "Company"},
                {"id": "Market", "text": "Market"},
                {"id": "Project", "text": "Project"},
                {"id": "Grant", "text": "Grant"}
            ]
        }
    ]

    
    return {
        "dataset": dataset,
        "stream": stream,
        "view_id": "blocks",
        "config": {
            "blocks": blocks,
            "global_css": "/*.prodigy-meta {display: none} */ .prodigy-container {max-width: 100%}"

        }
    }