hBERTv1_new_pretrain_48_emb_com_sst2

This model is a fine-tuned version of gokuls/bert_12_layer_model_v1_complete_training_new_emb_compress_48 on the GLUE SST2 dataset. It achieves the following results on the evaluation set:

• Loss: 0.4656
• Accuracy: 0.7947

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: 4e-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
0.4711	1.0	527	0.4730	0.7787
0.2942	2.0	1054	0.5417	0.7993
0.228	3.0	1581	0.5140	0.7901
0.1897	4.0	2108	0.4951	0.7947
0.1608	5.0	2635	0.4656	0.7947
0.1422	6.0	3162	0.6786	0.7741
0.1244	7.0	3689	0.6886	0.8028
0.1112	8.0	4216	0.7387	0.7821
0.0985	9.0	4743	0.7169	0.7936
0.0883	10.0	5270	0.7695	0.7959

Framework versions

• Transformers 4.30.2
• Pytorch 1.14.0a0+410ce96
• Datasets 2.12.0
• Tokenizers 0.13.3