tiny-mlm-snli-target-glue-stsb

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

• Loss: 0.9081
• Pearson: 0.8117
• Spearmanr: 0.8078

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	Pearson	Spearmanr
3.2807	2.78	500	1.2841	0.6694	0.6877
1.1742	5.56	1000	1.0700	0.7574	0.7698
0.8491	8.33	1500	0.9843	0.7918	0.7951
0.6909	11.11	2000	0.8178	0.8113	0.8070
0.5773	13.89	2500	0.8630	0.8127	0.8090
0.5095	16.67	3000	0.9598	0.8122	0.8106
0.4454	19.44	3500	0.8550	0.8159	0.8118
0.3945	22.22	4000	0.9033	0.8122	0.8080
0.357	25.0	4500	0.9651	0.8129	0.8091
0.3329	27.78	5000	0.9081	0.8117	0.8078

Framework versions

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