nbbert

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

• Accuracy: 0.9305
• Precision: 0.9342
• Recall: 0.9305
• F1: 0.9305
• Loss: 0.4443

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: 5e-05
• train_batch_size: 4
• eval_batch_size: 4
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 8
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 20
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Accuracy	Precision	Recall	F1	Validation Loss
No log	0.9412	8	0.4361	0.6823	0.4361	0.3171	0.8924
No log	2.0	17	0.8851	0.8758	0.8851	0.8748	0.4652
No log	2.9412	25	0.8281	0.8333	0.8281	0.8204	0.5819
No log	4.0	34	0.8759	0.8922	0.8759	0.8749	0.4312
No log	4.9412	42	0.8550	0.8762	0.8550	0.8548	0.5312
No log	6.0	51	0.8944	0.8940	0.8944	0.8941	0.3318
No log	6.9412	59	0.9209	0.9255	0.9209	0.9210	0.3824
No log	8.0	68	0.9213	0.9282	0.9213	0.9219	0.4385
No log	8.9412	76	0.9205	0.9226	0.9205	0.9205	0.3830
No log	10.0	85	0.9249	0.9309	0.9249	0.9252	0.4137
No log	10.9412	93	0.9269	0.9310	0.9269	0.9270	0.4014
No log	12.0	102	0.9293	0.9321	0.9293	0.9293	0.3923
No log	12.9412	110	0.9277	0.9320	0.9277	0.9278	0.4565
No log	14.0	119	0.9305	0.9342	0.9305	0.9305	0.4166
No log	14.9412	127	0.9281	0.9325	0.9281	0.9282	0.4512
No log	16.0	136	0.9297	0.9336	0.9297	0.9298	0.4465
No log	16.9412	144	0.9273	0.9318	0.9273	0.9274	0.4624
No log	18.0	153	0.9277	0.9321	0.9277	0.9278	0.4593
No log	18.8235	160	0.9305	0.9342	0.9305	0.9305	0.4443

Framework versions

• Transformers 4.44.2
• Pytorch 2.4.1+cu121
• Datasets 3.0.2
• Tokenizers 0.19.1