snippet to read .jsonl from Doccano NER annotator and converting into spacy v3 format. filter spans is optional, uncomment if you do not want overlapping span

doccano_jsonl_spacy3_read_converter.py

custom_data_path = 'path_to_your_ner_dataset.jsonl'
import json
 
with open(custom_data_path, 'r', encoding="utf8") as json_file:
    json_list = list(json_file)

dataset = [json.loads(jline) for jline in json_list]
print(f"number of examples : {len(dataset)}")


for training_example in tqdm(dataset): 
    text = training_example['data']
    labels = training_example['label']
    labels = [tuple(lbl) for lbl in labels]
    doc = nlp.make_doc(text) 
    ents = []
    for start, end, label in labels:
        span = doc.char_span(start, end, label=label, alignment_mode="contract")
        if span is None:
            print("Skipping entity")
        else:
            ents.append(span)
#     filtered_ents = filter_spans(ents)
#     doc.ents = filtered_ents 
    doc.ents = ents 
    doc_bin.add(doc)

doc_bin.to_disk("training_data.spacy") # save the docbin object