tiny-mlm-glue-qqp-target-glue-stsb

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

• Loss: 0.9234
• Pearson: 0.8132
• Spearmanr: 0.8116

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
2.9883	2.78	500	1.1659	0.7141	0.7498
0.9795	5.56	1000	1.0600	0.7790	0.8006
0.753	8.33	1500	0.9585	0.8042	0.8166
0.6208	11.11	2000	0.8495	0.8153	0.8188
0.5239	13.89	2500	0.8834	0.8149	0.8174
0.4691	16.67	3000	0.9556	0.8160	0.8195
0.4148	19.44	3500	0.8703	0.8180	0.8178
0.3779	22.22	4000	0.9027	0.8179	0.8177
0.3446	25.0	4500	0.9613	0.8191	0.8194
0.3215	27.78	5000	0.9470	0.8162	0.8160
0.3034	30.56	5500	0.9345	0.8161	0.8158
0.28	33.33	6000	0.9234	0.8132	0.8116

Framework versions

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