Instructions to use fktime/dbert_ai4p with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use fktime/dbert_ai4p with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("token-classification", model="fktime/dbert_ai4p")# pip install -U transformers accelerate # Load model directly from transformers import AutoTokenizer, AutoModelForTokenClassification tokenizer = AutoTokenizer.from_pretrained("fktime/dbert_ai4p") model = AutoModelForTokenClassification.from_pretrained("fktime/dbert_ai4p", device_map="auto") - Notebooks
- Google Colab
- Kaggle
Update README.md
Browse files
README.md
CHANGED
|
@@ -1 +1,6 @@
|
|
| 1 |
-
Overall F1 score of 92.67% and precision, recall, and accuracy metrics at 91.86%, 93.49%, and 97.49%, respectively
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
Overall F1 score of 92.67% and precision, recall, and accuracy metrics at 91.86%, 93.49%, and 97.49%, respectively.
|
| 2 |
+
|
| 3 |
+
High-Performing Entities: entities like Account Name, Account Number, Age, Company Name, and Email have achieved F1 scores above 95%.
|
| 4 |
+
Entities That Need Improvement: struggled with specific entities like Currency Name (27.3% F1) and IP (0.78% F1).
|
| 5 |
+
Numeric Entities: generally performed well with numeric entities like SSN (99.3% F1), Phone Number (97.5% F1), and Credit Card Issuer (98.1% F1).
|
| 6 |
+
Legal Entities: Job Title and Job Type, which might relate to the legal domain, achieved high F1 scores (around 97%). But need more entities!
|