du33169/t5-base-finetuned-GLUE-MRPC

MRPC

This model is a fine-tuned version of google-t5/t5-base on the GLUE MRPC dataset. It achieves the following results on the evaluation set:

• Loss: 0.5629
• Accuracy: 0.8971
• F1: 0.9268
• Combined Score: 0.9119

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: 32
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 10.0

Training results

Training Loss	Epoch	Step	Accuracy	Combined Score	F1	Validation Loss
No log	1.0	115	0.7108	0.7671	0.8234	0.5476
No log	2.0	230	0.8701	0.8901	0.9100	0.3523
No log	3.0	345	0.8725	0.8924	0.9122	0.3624
No log	4.0	460	0.8775	0.8949	0.9123	0.3646
0.3744	5.0	575	0.8946	0.9099	0.9252	0.4054
0.3744	6.0	690	0.8897	0.9057	0.9217	0.4624
0.3744	7.0	805	0.5530	0.8873	0.9212	0.9042
0.3744	8.0	920	0.5405	0.8897	0.9220	0.9059
0.0877	9.0	1035	0.5629	0.8971	0.9268	0.9119
0.0877	10.0	1150	0.5856	0.8922	0.9241	0.9081

Framework versions

• Transformers 4.43.3
• Pytorch 1.11.0+cu113
• Datasets 2.20.0
• Tokenizers 0.19.1