tiny-mlm-glue-cola-target-glue-stsb

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

• Loss: 0.9450
• Pearson: 0.8092
• Spearmanr: 0.8062

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
3.2852	2.78	500	1.2419	0.6797	0.6892
1.1811	5.56	1000	1.0907	0.7559	0.7738
0.8412	8.33	1500	0.9899	0.7947	0.8048
0.6828	11.11	2000	0.8542	0.8105	0.8119
0.5756	13.89	2500	0.8881	0.8122	0.8119
0.5049	16.67	3000	0.9491	0.8120	0.8121
0.4405	19.44	3500	0.8854	0.8131	0.8107
0.399	22.22	4000	0.9128	0.8112	0.8077
0.3565	25.0	4500	0.9623	0.8119	0.8096
0.3345	27.78	5000	0.9450	0.8092	0.8062

Framework versions

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