gerskill-gbert

This model is a fine-tuned version of deepset/gbert-base on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.1095
• Hard: {'precision': 0.7445544554455445, 'recall': 0.8245614035087719, 'f1': 0.7825182101977106, 'number': 456}
• Soft: {'precision': 0.7272727272727273, 'recall': 0.7804878048780488, 'f1': 0.7529411764705882, 'number': 82}
• Overall Precision: 0.7420
• Overall Recall: 0.8178
• Overall F1: 0.7781
• Overall Accuracy: 0.9635

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 2e-05
• train_batch_size: 32
• eval_batch_size: 32
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 5

Training results

Training Loss	Epoch	Step	Validation Loss	Hard	Soft	Overall Precision	Overall Recall	Overall F1	Overall Accuracy
No log	1.0	178	0.1255	{'precision': 0.583756345177665, 'recall': 0.756578947368421, 'f1': 0.659025787965616, 'number': 456}	{'precision': 0.5425531914893617, 'recall': 0.6219512195121951, 'f1': 0.5795454545454546, 'number': 82}	0.5781	0.7361	0.6476	0.9515
No log	2.0	356	0.1071	{'precision': 0.6994219653179191, 'recall': 0.7960526315789473, 'f1': 0.7446153846153847, 'number': 456}	{'precision': 0.6129032258064516, 'recall': 0.6951219512195121, 'f1': 0.6514285714285714, 'number': 82}	0.6863	0.7807	0.7304	0.9585
0.1562	3.0	534	0.0990	{'precision': 0.7150943396226415, 'recall': 0.831140350877193, 'f1': 0.7687626774847871, 'number': 456}	{'precision': 0.6777777777777778, 'recall': 0.7439024390243902, 'f1': 0.7093023255813954, 'number': 82}	0.7097	0.8178	0.7599	0.9621
0.1562	4.0	712	0.1072	{'precision': 0.7258687258687259, 'recall': 0.8245614035087719, 'f1': 0.7720739219712526, 'number': 456}	{'precision': 0.7222222222222222, 'recall': 0.7926829268292683, 'f1': 0.7558139534883721, 'number': 82}	0.7253	0.8197	0.7696	0.9628
0.1562	5.0	890	0.1095	{'precision': 0.7445544554455445, 'recall': 0.8245614035087719, 'f1': 0.7825182101977106, 'number': 456}	{'precision': 0.7272727272727273, 'recall': 0.7804878048780488, 'f1': 0.7529411764705882, 'number': 82}	0.7420	0.8178	0.7781	0.9635

Framework versions

• Transformers 4.38.1
• Pytorch 2.1.2+cu121
• Datasets 2.18.0
• Tokenizers 0.15.2