bert-base-uncased-sst2-unstructured-sparsity-80

This model is a fine-tuned version of bert-base-uncased on the GLUE SST2 dataset. The sparsity on linear layers is 80%.

It achieves the following results on the evaluation set:

• eval_loss: 0.4133
• eval_accuracy: 0.9128
• eval_runtime: 31.5327
• eval_samples_per_second: 27.654
• eval_steps_per_second: 3.457

Model description

• eval config： max_seq_length 128

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Framework versions

• Transformers 4.26.0
• Pytorch 1.13.1+cu116
• Datasets 2.8.0
• Tokenizers 0.13.2