Instructions to use THemidli/applied-ner-stage3-bert-tiny with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use THemidli/applied-ner-stage3-bert-tiny with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("token-classification", model="THemidli/applied-ner-stage3-bert-tiny")# Load model directly from transformers import AutoTokenizer, AutoModelForTokenClassification tokenizer = AutoTokenizer.from_pretrained("THemidli/applied-ner-stage3-bert-tiny") model = AutoModelForTokenClassification.from_pretrained("THemidli/applied-ner-stage3-bert-tiny", device_map="auto") - Notebooks
- Google Colab
- Kaggle
| { | |
| "stage": 3, | |
| "model_path": "models/stage3-bert-tiny-ner", | |
| "records": 40, | |
| "patterns": 10, | |
| "overall": { | |
| "true_positive": 50.0, | |
| "gold_spans": 98.0, | |
| "predicted_spans": 111.0, | |
| "precision": 0.45045045045045046, | |
| "recall": 0.5102040816326531, | |
| "f1": 0.4784688995215311 | |
| }, | |
| "error_counts": { | |
| "boundary": 18, | |
| "false_positive": 22, | |
| "missed_entity": 9, | |
| "wrong_label": 21 | |
| }, | |
| "label_confusions": { | |
| "AMOUNT->PRODUCT": 2, | |
| "JOB->ORGANIZATION": 2, | |
| "LOCATION->ORGANIZATION": 2, | |
| "ORGANIZATION->PRODUCT": 1, | |
| "ORGANIZATION->WORKOFART": 1, | |
| "PERSON->WORKOFART": 1, | |
| "PRODUCT->ORGANIZATION": 5, | |
| "PRODUCT->PERSON": 1, | |
| "PRODUCT->WORKOFART": 1, | |
| "TIMEDATE->AMOUNT": 1, | |
| "TIMEDATE->ORGANIZATION": 1, | |
| "WORKOFART->ORGANIZATION": 2, | |
| "WORKOFART->PERSON": 1 | |
| }, | |
| "by_pattern": { | |
| "amount_boundary": { | |
| "records": 4, | |
| "metrics": { | |
| "true_positive": 4.0, | |
| "gold_spans": 8.0, | |
| "predicted_spans": 9.0, | |
| "precision": 0.4444444444444444, | |
| "recall": 0.5, | |
| "f1": 0.47058823529411764 | |
| }, | |
| "error_counts": { | |
| "boundary": 4, | |
| "false_positive": 1 | |
| }, | |
| "records_with_errors": 2 | |
| }, | |
| "company_product": { | |
| "records": 4, | |
| "metrics": { | |
| "true_positive": 1.0, | |
| "gold_spans": 10.0, | |
| "predicted_spans": 9.0, | |
| "precision": 0.1111111111111111, | |
| "recall": 0.1, | |
| "f1": 0.10526315789473685 | |
| }, | |
| "error_counts": { | |
| "boundary": 2, | |
| "false_positive": 2, | |
| "missed_entity": 3, | |
| "wrong_label": 4 | |
| }, | |
| "records_with_errors": 4 | |
| }, | |
| "institution_context": { | |
| "records": 4, | |
| "metrics": { | |
| "true_positive": 5.0, | |
| "gold_spans": 9.0, | |
| "predicted_spans": 12.0, | |
| "precision": 0.4166666666666667, | |
| "recall": 0.5555555555555556, | |
| "f1": 0.4761904761904762 | |
| }, | |
| "error_counts": { | |
| "boundary": 1, | |
| "false_positive": 4, | |
| "missed_entity": 1, | |
| "wrong_label": 2 | |
| }, | |
| "records_with_errors": 4 | |
| }, | |
| "job_boundary": { | |
| "records": 4, | |
| "metrics": { | |
| "true_positive": 6.0, | |
| "gold_spans": 13.0, | |
| "predicted_spans": 15.0, | |
| "precision": 0.4, | |
| "recall": 0.46153846153846156, | |
| "f1": 0.42857142857142855 | |
| }, | |
| "error_counts": { | |
| "boundary": 3, | |
| "false_positive": 3, | |
| "missed_entity": 1, | |
| "wrong_label": 3 | |
| }, | |
| "records_with_errors": 4 | |
| }, | |
| "namedness_repetition": { | |
| "records": 4, | |
| "metrics": { | |
| "true_positive": 7.0, | |
| "gold_spans": 9.0, | |
| "predicted_spans": 10.0, | |
| "precision": 0.7, | |
| "recall": 0.7777777777777778, | |
| "f1": 0.7368421052631577 | |
| }, | |
| "error_counts": { | |
| "false_positive": 1, | |
| "wrong_label": 2 | |
| }, | |
| "records_with_errors": 3 | |
| }, | |
| "person_alias_handle": { | |
| "records": 4, | |
| "metrics": { | |
| "true_positive": 10.0, | |
| "gold_spans": 12.0, | |
| "predicted_spans": 12.0, | |
| "precision": 0.8333333333333334, | |
| "recall": 0.8333333333333334, | |
| "f1": 0.8333333333333334 | |
| }, | |
| "error_counts": { | |
| "false_positive": 1, | |
| "missed_entity": 1, | |
| "wrong_label": 1 | |
| }, | |
| "records_with_errors": 2 | |
| }, | |
| "product_work": { | |
| "records": 4, | |
| "metrics": { | |
| "true_positive": 2.0, | |
| "gold_spans": 9.0, | |
| "predicted_spans": 12.0, | |
| "precision": 0.16666666666666666, | |
| "recall": 0.2222222222222222, | |
| "f1": 0.1904761904761905 | |
| }, | |
| "error_counts": { | |
| "boundary": 1, | |
| "false_positive": 4, | |
| "missed_entity": 1, | |
| "wrong_label": 5 | |
| }, | |
| "records_with_errors": 4 | |
| }, | |
| "software_namedness": { | |
| "records": 4, | |
| "metrics": { | |
| "true_positive": 4.0, | |
| "gold_spans": 10.0, | |
| "predicted_spans": 11.0, | |
| "precision": 0.36363636363636365, | |
| "recall": 0.4, | |
| "f1": 0.380952380952381 | |
| }, | |
| "error_counts": { | |
| "boundary": 2, | |
| "false_positive": 2, | |
| "missed_entity": 1, | |
| "wrong_label": 3 | |
| }, | |
| "records_with_errors": 3 | |
| }, | |
| "timedate_boundary": { | |
| "records": 4, | |
| "metrics": { | |
| "true_positive": 8.0, | |
| "gold_spans": 10.0, | |
| "predicted_spans": 13.0, | |
| "precision": 0.6153846153846154, | |
| "recall": 0.8, | |
| "f1": 0.6956521739130435 | |
| }, | |
| "error_counts": { | |
| "boundary": 1, | |
| "false_positive": 3, | |
| "wrong_label": 1 | |
| }, | |
| "records_with_errors": 2 | |
| }, | |
| "work_title_boundary": { | |
| "records": 4, | |
| "metrics": { | |
| "true_positive": 3.0, | |
| "gold_spans": 8.0, | |
| "predicted_spans": 8.0, | |
| "precision": 0.375, | |
| "recall": 0.375, | |
| "f1": 0.375 | |
| }, | |
| "error_counts": { | |
| "boundary": 4, | |
| "false_positive": 1, | |
| "missed_entity": 1 | |
| }, | |
| "records_with_errors": 3 | |
| } | |
| }, | |
| "method": "Forty fresh, manually gold-labeled policy probes; exact character-span comparison; errors classified by deterministic overlap rules and then manually reviewed." | |
| } | |