mobilebert_sa_GLUE_Experiment_logit_kd_mnli_256

This model is a fine-tuned version of google/mobilebert-uncased on the GLUE MNLI dataset. It achieves the following results on the evaluation set:

• Loss: 1.2282
• Accuracy: 0.6120

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: 128
• eval_batch_size: 128
• 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

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
1.6433	1.0	3068	1.4078	0.5457
1.4683	2.0	6136	1.3590	0.5658
1.4077	3.0	9204	1.3106	0.5772
1.3591	4.0	12272	1.2971	0.5904
1.3213	5.0	15340	1.2764	0.5957
1.2849	6.0	18408	1.2562	0.6029
1.2475	7.0	21476	1.2524	0.6038
1.2073	8.0	24544	1.2384	0.6066
1.1713	9.0	27612	1.2377	0.6109
1.1371	10.0	30680	1.2228	0.6077
1.1069	11.0	33748	1.2126	0.6196
1.0775	12.0	36816	1.2232	0.6271
1.0491	13.0	39884	1.2440	0.6110
1.0228	14.0	42952	1.2741	0.6079
0.9977	15.0	46020	1.2448	0.6158
0.974	16.0	49088	1.3261	0.6206

Framework versions

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