bert_imdb

This model is a fine-tuned version of google-bert/bert-base-uncased on imdb dataset. It achieves the following results on the evaluation set:

• Loss: 0.3119
• Accuracy: 0.9403
• Recall: 0.9430
• Precision: 0.9379

To acccess my finetuning tutorial you can check the following repository.

Comparison with SOTA:

• DistilBERT 66M: 92.82
• BERT-base + ITPT: 95.63
• BERT-large: 95.49

Reference: Paperswithcode

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: 2e-05
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• num_epochs: 4

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Recall	Precision
0.2099	1.0	1563	0.2456	0.9102	0.8481	0.9683
0.1379	2.0	3126	0.2443	0.9274	0.8911	0.9608
0.0752	3.0	4689	0.2845	0.9391	0.9509	0.9290
0.0352	4.0	6252	0.3119	0.9403	0.9430	0.9379

Framework versions

• Transformers 4.47.1
• Pytorch 2.5.1+cu121
• Datasets 3.2.0
• Tokenizers 0.21.0