distilbert_sa_GLUE_Experiment_logit_kd_cola

This model is a fine-tuned version of distilbert-base-uncased on the GLUE COLA dataset. It achieves the following results on the evaluation set:

• Loss: 0.6741
• Matthews Correlation: -0.0207

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: 5e-05
• train_batch_size: 256
• eval_batch_size: 256
• seed: 10
• distributed_type: multi-GPU
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 50
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Matthews Correlation
0.814	1.0	34	0.6851	0.0
0.7923	2.0	68	0.6741	-0.0207
0.7521	3.0	102	0.7281	0.0931
0.6713	4.0	136	0.6815	0.0434
0.6052	5.0	170	0.7829	0.1374
0.5654	6.0	204	0.7213	0.1027
0.5296	7.0	238	0.8135	0.0702

Framework versions

• Transformers 4.26.0
• Pytorch 1.14.0a0+410ce96
• Datasets 2.9.0
• Tokenizers 0.13.2