BERT_ST_DA_100_v2

This model is a fine-tuned version of google-bert/bert-base-uncased on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.2371
• Precision: 0.9457
• Recall: 0.9480
• F1: 0.9469
• Accuracy: 0.9446

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: 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	Precision	Recall	F1	Accuracy
No log	1.0	59	0.3489	0.9065	0.9194	0.9129	0.9085
No log	2.0	118	0.2883	0.9190	0.9267	0.9228	0.9180
No log	3.0	177	0.2505	0.9322	0.9403	0.9362	0.9330
No log	4.0	236	0.2300	0.9384	0.9446	0.9415	0.9384
No log	5.0	295	0.2305	0.9397	0.9435	0.9416	0.9386
No log	6.0	354	0.2332	0.9443	0.9482	0.9462	0.9438
No log	7.0	413	0.2341	0.9433	0.9468	0.9450	0.9429
No log	8.0	472	0.2364	0.9441	0.9474	0.9457	0.9430
0.1814	9.0	531	0.2339	0.9457	0.9472	0.9465	0.9439
0.1814	10.0	590	0.2371	0.9457	0.9480	0.9469	0.9446

Framework versions

• Transformers 4.44.2
• Pytorch 2.4.0+cu121
• Datasets 2.21.0
• Tokenizers 0.19.1