app.py

import gradio as gr
import numpy as np
from huggingface_hub import hf_hub_download
import os


def predict_label(text):
    
    ip = text.split()
    ip_len = [len(ip)]
    
    span_scores = extract_spannet_scores(span_model,ip,ip_len, pos_col=1, task_col=2)
    span_pooled_scores = pool_span_scores(span_scores, ip_len)

    msa_span_scores = extract_spannet_scores(msa_span_model,ip,ip_len, pos_col=1, task_col=2, pos='not none')
    msa_pooled_scores = pool_span_scores(msa_span_scores, ip_len)

    ensemble_span_scores = [score for scores in [span_scores, msa_span_scores] for score in scores]
    ensemble_pooled_scores = pool_span_scores(ensemble_span_scores, ip_len)
    
    ent_scores = extract_ent_scores(entity_model,ip,ensemble_pooled_scores, pos_col=1, task_col=2)
    combined_sequences, ent_pred_tags = pool_ent_scores(ent_scores, ip_len)
    
    return combined_sequences


if __name__ == '__main__':
    
    space_key = os.environ.get('key')
    filenames = ['network.py', 'layers.py', 'utils.py',
                 'representation.py', 'predict.py', 'validate.py']
    
    for file in filenames:
        hf_hub_download('nehalelkaref/stagedNER',
                    filename=file,
                    local_dir='src',
                    token=space_key)
    
    from src.predict import extract_spannet_scores,extract_ent_scores,pool_span_scores,pool_ent_scores
    from src.network import SpanNet, EntNet
    from src.validate import entities_from_token_classes
    
    span_path = 'models/span.model'
    msa_span_path = 'models/msa.best.model'
    entity_path= 'models/entity.msa.model'
    
    span_model = SpanNet.load_model(span_path)
    msa_span_model = SpanNet.load_model(msa_span_path)
    entity_model = EntNet.load_model(entity_path)
    # iface= gr.Base(primary_hue="green")
    iface = gr.Interface(fn=predict_label, inputs="text", outputs="text")
    iface.launch(show_api=False)