visobert-human-tl-seg-seed-6969

This model is a fine-tuned version of uitnlp/visobert on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.3765
• Accuracy: 0.8626
• Precision: 0.7099
• Recall: 0.5885
• F1: 0.6157

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
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 500
• num_epochs: 50

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision	Recall	F1
No log	1.0	346	0.4179	0.8439	0.6639	0.4685	0.4911
0.5071	2.0	692	0.3976	0.8582	0.7045	0.5327	0.5664
0.3754	3.0	1038	0.3962	0.8589	0.7381	0.5336	0.5745
0.3754	4.0	1384	0.3867	0.8574	0.7182	0.5499	0.5811
0.3711	5.0	1730	0.3849	0.8615	0.7316	0.5551	0.5903
0.3634	6.0	2076	0.3894	0.8589	0.7282	0.5462	0.5785
0.3634	7.0	2422	0.3789	0.8608	0.7043	0.5474	0.5898
0.3631	8.0	2768	0.3763	0.8623	0.7172	0.5535	0.5932
0.3551	9.0	3114	0.3728	0.8608	0.6908	0.5678	0.6064
0.3551	10.0	3460	0.3885	0.8585	0.7107	0.5637	0.5859
0.3613	11.0	3806	0.3747	0.8612	0.7077	0.5534	0.5947
0.3518	12.0	4152	0.3765	0.8626	0.7099	0.5885	0.6157
0.3518	13.0	4498	0.3746	0.8619	0.7195	0.5563	0.5935
0.354	14.0	4844	0.3716	0.8604	0.7016	0.5562	0.5977
0.3483	15.0	5190	0.3721	0.8612	0.7102	0.5588	0.5985
0.3516	16.0	5536	0.3739	0.8597	0.7213	0.5375	0.5787
0.3516	17.0	5882	0.3703	0.8597	0.6851	0.5854	0.6137

Framework versions

• Transformers 4.51.1
• Pytorch 2.5.1+cu124
• Datasets 3.5.0
• Tokenizers 0.21.0