Modelpublic

gokulsrinivasagan/tinybert_base_train_book_ent_15p_s_init_kd_complete_qqp

sourceHugging Faceapache-2.0updated 1y agoView on Hugging Face

0likes6downloads

Model Card

tinybertbasetrainbookent15psinitkdcompleteqqp

This model is a fine-tuned version of gokulsrinivasagan/tinybert_base_train_book_ent_15p_s_init_kd_complete on the GLUE QQP dataset. It achieves the following results on the evaluation set:

—Loss: 0.2906
—Accuracy: 0.8756
—F1: 0.8286
—Combined Score: 0.8521

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
—trainbatchsize: 256
—evalbatchsize: 256
—seed: 10
—optimizer: Use adamwtorch with betas=(0.9,0.999) and epsilon=1e-08 and optimizerargs=No additional optimizer arguments
—lrschedulertype: linear
—num_epochs: 50

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1	Combined Score
0.4076	1.0	1422	0.3414	0.8449	0.7806	0.8128
0.3116	2.0	2844	0.3196	0.8569	0.8224	0.8397
0.265	3.0	4266	0.2982	0.8736	0.8339	0.8537
0.2293	4.0	5688	0.2906	0.8756	0.8286	0.8521
0.1988	5.0	7110	0.3052	0.8808	0.8401	0.8605
0.172	6.0	8532	0.2961	0.8834	0.8451	0.8642
0.1498	7.0	9954	0.3296	0.8802	0.8434	0.8618
0.1299	8.0	11376	0.3442	0.8872	0.8495	0.8684
0.1142	9.0	12798	0.3699	0.8851	0.8474	0.8662

Framework versions

—Transformers 4.51.2
—Pytorch 2.6.0+cu126
—Datasets 3.5.0
—Tokenizers 0.21.1