tiny-mlm-glue-mrpc-target-glue-stsb

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

• Loss: 0.9256
• Pearson: 0.8144
• Spearmanr: 0.8124

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
• num_epochs: 200

Training results

Training Loss	Epoch	Step	Validation Loss	Pearson	Spearmanr
2.9549	2.78	500	1.1776	0.7187	0.7620
0.9497	5.56	1000	1.1286	0.7777	0.8045
0.7413	8.33	1500	1.0286	0.8024	0.8197
0.6158	11.11	2000	0.8676	0.8181	0.8232
0.5192	13.89	2500	0.9099	0.8164	0.8193
0.4693	16.67	3000	1.0002	0.8169	0.8212
0.4144	19.44	3500	0.8905	0.8185	0.8185
0.3768	22.22	4000	0.9210	0.8193	0.8191
0.3443	25.0	4500	0.9723	0.8201	0.8206
0.3212	27.78	5000	0.9919	0.8157	0.8162
0.3032	30.56	5500	0.9692	0.8164	0.8162
0.2797	33.33	6000	0.9256	0.8144	0.8124

Framework versions

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