router-mmBERT-small-text-only-v3

This model is a fine-tuned version of jhu-clsp/mmBERT-small on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.5264
• Accuracy: 0.7443
• Precision: 0.7422
• Recall: 0.7443
• F1: 0.7282

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: 16
• eval_batch_size: 32
• seed: 42
• optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: cosine
• num_epochs: 2

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision	Recall	F1
No log	0	0	1.5114	0.4432	0.4804	0.4432	0.4537
0.7931	0.2273	20	0.9312	0.6648	0.6848	0.6648	0.5784
0.7752	0.4545	40	0.9223	0.4886	0.7377	0.4886	0.4320
0.5251	0.6818	60	0.5346	0.6989	0.6940	0.6989	0.6659
0.5975	0.9091	80	0.5975	0.6534	0.7276	0.6534	0.6573
0.551	1.1364	100	0.5680	0.7273	0.7225	0.7273	0.7090
0.5093	1.3636	120	0.5872	0.7045	0.6952	0.7045	0.6951
0.5315	1.5909	140	0.5398	0.75	0.7593	0.75	0.7265
0.5231	1.8182	160	0.5264	0.7443	0.7422	0.7443	0.7282

Framework versions

• Transformers 4.57.1
• Pytorch 2.8.0+cu128
• Datasets 4.2.0
• Tokenizers 0.22.1