curated-gender-equality-ordinal-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: 0.1956
• Accuracy: 0.9603
• F1: 0.9485
• Precision: 0.9288
• Recall: 0.9690

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: 24
• eval_batch_size: 24
• 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
0.653	1.0	342	0.4418	0.7789	0.7700	0.6341	0.9800
0.3746	2.0	684	0.3035	0.9164	0.8967	0.8403	0.9613
0.2861	3.0	1026	0.2572	0.9425	0.9268	0.8919	0.9645
0.2451	4.0	1368	0.2372	0.9449	0.9297	0.8968	0.9651
0.2205	5.0	1710	0.2252	0.9483	0.9340	0.9020	0.9684
0.2019	6.0	2052	0.2159	0.9488	0.9347	0.9021	0.9697
0.1881	7.0	2394	0.2091	0.9520	0.9384	0.9102	0.9684
0.1762	8.0	2736	0.2035	0.9551	0.9422	0.9174	0.9684
0.1699	9.0	3078	0.2052	0.9534	0.9402	0.9125	0.9697
0.1572	10.0	3420	0.1975	0.9576	0.9452	0.9226	0.9690
0.1502	11.0	3762	0.1997	0.9576	0.9452	0.9220	0.9697
0.1451	12.0	4104	0.1934	0.9600	0.9481	0.9293	0.9677
0.1406	13.0	4446	0.1994	0.9569	0.9444	0.9198	0.9703
0.1375	14.0	4788	0.1912	0.9612	0.9497	0.9311	0.9690
0.1321	15.0	5130	0.1943	0.9600	0.9482	0.9288	0.9684
0.1275	16.0	5472	0.1974	0.9603	0.9485	0.9288	0.9690
0.1269	17.0	5814	0.1928	0.9610	0.9493	0.9316	0.9677
0.1264	18.0	6156	0.1924	0.9617	0.9502	0.9334	0.9677
0.1244	19.0	6498	0.1951	0.9605	0.9488	0.9294	0.9690
0.1195	20.0	6840	0.1956	0.9603	0.9485	0.9288	0.9690

Framework versions

• Transformers 4.42.4
• Pytorch 2.4.0+cu121
• Datasets 2.21.0
• Tokenizers 0.19.1