| --- | |
| license: apache-2.0 | |
| base_model: albert/albert-base-v2 | |
| tags: | |
| - generated_from_trainer | |
| metrics: | |
| - accuracy | |
| - f1 | |
| - precision | |
| - recall | |
| model-index: | |
| - name: classify-clickbait-titll | |
| results: [] | |
| Identify Clickbait Articles | |
| This model is a fine-tuned version of albert/albert-base-v2 on a synthetic dataset with 65% ISIN titles and 35% ISIN_null titles. | |
| ### Model description | |
| Built to identify ISIN vs ISIN_null titles. | |
| ### Intended uses & limitations | |
| Use it on any title to understand how the model is interpreting the title, whether it is ISIN or ISIN_null. | |
| Go ahead and try a few of your own. | |
| ### Training and evaluation data | |
| It achieves the following results on the evaluation set: | |
| Loss: 0.0173 | |
| Accuracy: 0.9951 | |
| F1: 0.9951 | |
| Precision: 0.9951 | |
| Recall: 0.9951 | |
| Accuracy Label ISIN: 0.95 | |
| Accuracy Label ISIN_null: .095 | |
| Training procedure | |
| Training hyperparameters | |
| ### Training hyperparameters | |
| The following hyperparameters were used during training: | |
| - learning_rate: 2e-05 | |
| - train_batch_size: 16 | |
| - eval_batch_size: 16 | |
| - seed: 42 | |
| - gradient_accumulation_steps: 2 | |
| - total_train_batch_size: 32 | |
| - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 | |
| - lr_scheduler_type: linear | |
| - lr_scheduler_warmup_steps: 500 | |
| - num_epochs: 280 | |
| ### Framework versions | |
| Transformers 4.43.3 | |
| Datasets 2.20.0 | |
| Tokenizers 0.19.1 |