muhtasham/bert-small-finetuned-eoir-privacy-longer30

bert_uncased_L-4_H-512_A-8-finetuned-eoir_privacy-longer-finetuned-eoir_privacy-longer30

This model was trained from scratch on the eoir_privacy dataset. It achieves the following results on the evaluation set:

• Loss: 0.2076
• Accuracy: 0.9491
• F1: 0.8838

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: 8
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 16
• total_train_batch_size: 128
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1
No log	1.0	63	0.2446	0.9426	0.8671
No log	2.0	126	0.2306	0.9412	0.8629
No log	3.0	189	0.2637	0.9448	0.8679
No log	4.0	252	0.2375	0.9455	0.8758
No log	5.0	315	0.2423	0.9440	0.8687
No log	6.0	378	0.2571	0.9455	0.8676
No log	7.0	441	0.2040	0.9469	0.8799
0.0355	8.0	504	0.2096	0.9462	0.8784
0.0355	9.0	567	0.2094	0.9448	0.8736
0.0355	10.0	630	0.2076	0.9491	0.8838

Framework versions

• Transformers 4.21.1
• Pytorch 1.12.1+cu113
• Datasets 2.4.0
• Tokenizers 0.12.1