small-mlm-glue-mnli-target-glue-mrpc

This model is a fine-tuned version of muhtasham/small-mlm-glue-mnli on the None dataset. It achieves the following results on the evaluation set:

• Loss: 1.9598
• Accuracy: 0.7721
• F1: 0.8432

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: 3e-05
• train_batch_size: 32
• eval_batch_size: 32
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: constant
• training_steps: 5000

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1
0.3946	4.35	500	0.8309	0.7647	0.8476
0.0715	8.7	1000	1.3662	0.7647	0.8405
0.0285	13.04	1500	1.6659	0.7647	0.8405
0.0149	17.39	2000	1.8421	0.7696	0.8396
0.0158	21.74	2500	1.9587	0.7647	0.8426
0.0152	26.09	3000	2.0488	0.7672	0.848
0.0147	30.43	3500	1.9463	0.7770	0.8535
0.0096	34.78	4000	1.7938	0.7819	0.8529
0.0124	39.13	4500	1.8361	0.7868	0.8538
0.0152	43.48	5000	1.9598	0.7721	0.8432

Framework versions

• Transformers 4.26.0.dev0
• Pytorch 1.13.0+cu116
• Datasets 2.8.1.dev0
• Tokenizers 0.13.2