bert-base-cased_12112024T103442

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

• Loss: 0.4204
• F1: 0.8800
• Learning Rate: 0.0

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: 1e-05
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 32
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine
• lr_scheduler_warmup_steps: 600
• num_epochs: 20
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	F1	Rate
No log	0.9942	86	1.8033	0.0562	0.0000
No log	2.0	173	1.6866	0.2347	0.0000
No log	2.9942	259	1.5021	0.4551	0.0000
No log	4.0	346	1.2315	0.5317	0.0000
No log	4.9942	432	1.0796	0.5664	0.0000
1.4663	6.0	519	0.9279	0.6285	0.0000
1.4663	6.9942	605	0.8522	0.6722	1e-05
1.4663	8.0	692	0.7117	0.7331	0.0000
1.4663	8.9942	778	0.6128	0.7896	0.0000
1.4663	10.0	865	0.5323	0.8263	0.0000
1.4663	10.9942	951	0.5330	0.8196	0.0000
0.6158	12.0	1038	0.4660	0.8616	0.0000
0.6158	12.9942	1124	0.4204	0.8800	0.0000
0.6158	14.0	1211	0.4407	0.8770	0.0000
0.6158	14.9942	1297	0.4435	0.8780	0.0000
0.6158	16.0	1384	0.4412	0.8791	0.0000
0.6158	16.9942	1470	0.4424	0.8802	0.0000
0.1869	18.0	1557	0.4466	0.8809	5e-07
0.1869	18.9942	1643	0.4469	0.8795	1e-07
0.1869	19.8844	1720	0.4483	0.8798	0.0

Framework versions

• Transformers 4.44.2
• Pytorch 2.5.1+cu124
• Datasets 3.1.0
• Tokenizers 0.19.1