bert_tiny_lda_20_mrpc

This model is a fine-tuned version of gokulsrinivasagan/bert_tiny_lda_20 on the GLUE MRPC dataset. It achieves the following results on the evaluation set:

• Loss: 0.6247
• Accuracy: 0.6838
• F1: 0.8122
• Combined Score: 0.7480

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: 0.001
• 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: 30

Training results

Training Loss	Epoch	Step	Accuracy	Combined Score	F1	Validation Loss
0.6805	1.0	39	0.6838	0.7480	0.8122	0.6292
0.6386	2.0	78	0.6838	0.7480	0.8122	0.6304
0.6329	3.0	117	0.6838	0.7480	0.8122	0.6240
0.6373	4.0	156	0.6838	0.7480	0.8122	0.6240
0.6361	5.0	195	0.6254	0.6838	0.8122	0.7480
0.6372	6.0	234	0.6244	0.6838	0.8122	0.7480
0.6324	7.0	273	0.6249	0.6838	0.8122	0.7480
0.6344	8.0	312	0.6241	0.6838	0.8122	0.7480
0.631	9.0	351	0.6279	0.6838	0.8122	0.7480

Framework versions

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