bert_base_for_whole_train_result_Spam-Ham_farshad_half_1_2

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

• Loss: 0.0762
• Accuracy: 0.9875
• F1: 0.9878

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: 0.0001
• train_batch_size: 64
• eval_batch_size: 64
• seed: 42
• gradient_accumulation_steps: 64
• total_train_batch_size: 4096
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 1000
• num_epochs: 100

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1
0.5984	5.8501	50	0.3899	0.9014	0.9010
0.2022	11.7002	100	0.0748	0.9780	0.9785
0.0405	17.5503	150	0.0594	0.9806	0.9810
0.0148	23.4004	200	0.0317	0.9925	0.9927
0.0082	29.2505	250	0.0412	0.9910	0.9913
0.0056	35.1005	300	0.0396	0.9919	0.9922
0.0041	40.9506	350	0.0793	0.9823	0.9827
0.0026	46.8007	400	0.0444	0.9907	0.9910
0.0024	52.6508	450	0.0656	0.9867	0.9870
0.0016	58.5009	500	0.0474	0.9913	0.9916
0.001	64.3510	550	0.0595	0.9907	0.9910
0.0018	70.2011	600	0.0521	0.9907	0.9911
0.0014	76.0512	650	0.0544	0.9919	0.9921
0.0011	81.9013	700	0.0762	0.9875	0.9878

Framework versions

• Transformers 4.40.0
• Pytorch 2.4.1+cu121
• Datasets 3.1.0
• Tokenizers 0.19.1