bert_base_train_book_ent_40p_inv_stsb

This model is a fine-tuned version of gokulsrinivasagan/bert_base_train_book_ent_40p_inv on the GLUE STSB dataset. It achieves the following results on the evaluation set:

• Loss: 1.2772
• Pearson: 0.6656
• Spearmanr: 0.6556
• Combined Score: 0.6606

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: 5e-05
• train_batch_size: 256
• eval_batch_size: 256
• seed: 10
• optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• num_epochs: 50

Training results

Training Loss	Epoch	Step	Validation Loss	Pearson	Spearmanr	Combined Score
2.5604	1.0	23	2.5837	0.1641	0.1684	0.1663
1.6885	2.0	46	2.1112	0.4817	0.4723	0.4770
1.0625	3.0	69	1.5126	0.5978	0.6087	0.6032
0.7028	4.0	92	1.7524	0.6360	0.6605	0.6482
0.4906	5.0	115	1.4309	0.6410	0.6400	0.6405
0.4045	6.0	138	1.7941	0.6649	0.6729	0.6689
0.3196	7.0	161	1.4584	0.6809	0.6811	0.6810
0.2451	8.0	184	1.3930	0.6636	0.6588	0.6612
0.2024	9.0	207	1.3720	0.6676	0.6625	0.6650
0.206	10.0	230	1.3823	0.6622	0.6555	0.6589
0.1855	11.0	253	1.5228	0.6638	0.6598	0.6618
0.1549	12.0	276	1.3798	0.6693	0.6632	0.6663
0.1351	13.0	299	1.2772	0.6656	0.6556	0.6606
0.1358	14.0	322	1.4514	0.6670	0.6623	0.6647
0.1189	15.0	345	1.4984	0.6602	0.6537	0.6569
0.0982	16.0	368	1.2806	0.6705	0.6619	0.6662
0.0995	17.0	391	1.3375	0.6672	0.6589	0.6631
0.0896	18.0	414	1.3894	0.6575	0.6491	0.6533

Framework versions

• Transformers 4.46.3
• Pytorch 2.2.1+cu118
• Datasets 3.1.0
• Tokenizers 0.20.3