uae-climate-multi-classifier-weighted

This model is a fine-tuned version of alex-miller/ODABert on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 1.0936
• Accuracy: 0.9197
• F1: 0.7059
• Precision: 0.7
• Recall: 0.7119

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: 1e-06
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1	Precision	Recall
1.2281	1.0	246	1.4142	0.8670	0.0333	1.0	0.0169
1.1785	2.0	492	1.3474	0.8830	0.3014	0.7857	0.1864
1.0939	3.0	738	1.1641	0.8784	0.4421	0.5833	0.3559
0.9503	4.0	984	1.1304	0.9083	0.5455	0.8276	0.4068
0.8772	5.0	1230	1.1369	0.9197	0.6154	0.875	0.4746
1.0567	6.0	1476	0.9456	0.9151	0.6667	0.7115	0.6271
0.9089	7.0	1722	0.9752	0.9174	0.6727	0.7255	0.6271
0.9149	8.0	1968	1.1084	0.9197	0.6667	0.7609	0.5932
0.8285	9.0	2214	0.8487	0.9037	0.6818	0.6164	0.7627
0.8322	10.0	2460	1.1982	0.9174	0.6471	0.7674	0.5593
0.7955	11.0	2706	1.0507	0.9174	0.6667	0.7347	0.6102
0.7942	12.0	2952	0.9965	0.9197	0.7059	0.7	0.7119
0.7152	13.0	3198	1.0222	0.9197	0.7009	0.7069	0.6949
0.7015	14.0	3444	0.9579	0.9197	0.7244	0.6765	0.7797
0.5537	15.0	3690	1.0659	0.9220	0.7069	0.7193	0.6949
0.6379	16.0	3936	1.0921	0.9197	0.6957	0.7143	0.6780
0.5977	17.0	4182	1.0491	0.9220	0.7213	0.6984	0.7458
0.6792	18.0	4428	1.0872	0.9197	0.7059	0.7	0.7119
0.5509	19.0	4674	1.0763	0.9220	0.7167	0.7049	0.7288
0.579	20.0	4920	1.0936	0.9197	0.7059	0.7	0.7119

Framework versions

• Transformers 4.39.3
• Pytorch 2.2.2+cu121
• Datasets 2.18.0
• Tokenizers 0.15.2