distilbert_add_GLUE_Experiment_logit_kd_rte

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

• Loss: 0.4229
• Accuracy: 0.4729

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	Accuracy
0.4684	1.0	10	0.4310	0.4729
0.4214	2.0	20	0.4342	0.4729
0.4216	3.0	30	0.4264	0.4729
0.4197	4.0	40	0.4311	0.4729
0.425	5.0	50	0.4297	0.4729
0.4192	6.0	60	0.4260	0.4729
0.4182	7.0	70	0.4243	0.4729
0.4184	8.0	80	0.4246	0.4729
0.4201	9.0	90	0.4240	0.4729
0.417	10.0	100	0.4259	0.4729
0.419	11.0	110	0.4269	0.4729
0.4165	12.0	120	0.4249	0.4729
0.4116	13.0	130	0.4229	0.4729
0.3924	14.0	140	0.4916	0.4729
0.3783	15.0	150	0.4539	0.4874
0.3384	16.0	160	0.4581	0.4982
0.3202	17.0	170	0.5284	0.4765
0.3054	18.0	180	0.4884	0.5162

Framework versions

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