multilingual-e5-base_censor_v0.1

This model is a fine-tuned version of intfloat/multilingual-e5-base on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.5077
• Accuracy: 0.7695
• Precision: 0.7729
• Recall: 0.7695
• F1: 0.7708
• Roc Auc: 0.8424
• Per Class F1: [0.8104358705451601, 0.7061407261629732]

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: 2e-05
• train_batch_size: 64
• eval_batch_size: 64
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 500
• num_epochs: 3
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision	Recall	F1	Roc Auc	Per Class F1
0.5485	1.0	2795	0.5249	0.7349	0.7498	0.7349	0.7383	0.8169	[0.7723692804179954, 0.6827648980028912]
0.4871	2.0	5590	0.5059	0.7610	0.7667	0.7610	0.7629	0.8362	[0.8013980033834657, 0.7000926419808541]
0.4392	3.0	8385	0.5077	0.7695	0.7729	0.7695	0.7708	0.8424	[0.8104358705451601, 0.7061407261629732]

Framework versions

• Transformers 4.43.4
• Pytorch 2.3.1+cu121
• Datasets 2.20.0
• Tokenizers 0.19.1