bert_base_train_book_stsb

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

• Loss: 0.8312
• Pearson: 0.7987
• Spearmanr: 0.7962
• Combined Score: 0.7975

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
2.7053	1.0	23	3.0797	0.2078	0.1978	0.2028
1.6079	2.0	46	1.3635	0.6578	0.6708	0.6643
0.8744	3.0	69	1.0689	0.7391	0.7444	0.7418
0.6927	4.0	92	0.9753	0.7538	0.7525	0.7532
0.513	5.0	115	0.8901	0.7786	0.7776	0.7781
0.4145	6.0	138	0.9329	0.8023	0.8040	0.8032
0.3148	7.0	161	1.0895	0.7820	0.7806	0.7813
0.2547	8.0	184	0.9650	0.7741	0.7760	0.7750
0.2385	9.0	207	0.8644	0.7980	0.7967	0.7973
0.1984	10.0	230	0.8782	0.7990	0.8005	0.7998
0.165	11.0	253	0.9499	0.8044	0.8060	0.8052
0.1628	12.0	276	1.0373	0.7997	0.8026	0.8011
0.1328	13.0	299	0.9285	0.7979	0.8006	0.7992
0.1235	14.0	322	0.8312	0.7987	0.7962	0.7975
0.1061	15.0	345	0.8854	0.8044	0.8034	0.8039
0.1095	16.0	368	0.9605	0.7908	0.7908	0.7908
0.1052	17.0	391	0.8527	0.7919	0.7897	0.7908
0.0939	18.0	414	0.8989	0.7928	0.7926	0.7927
0.0898	19.0	437	1.0319	0.7994	0.8021	0.8008

Framework versions

• Transformers 4.46.3
• Pytorch 2.2.1+cu118
• Datasets 2.17.0
• Tokenizers 0.20.3