N1CKNGUYEN/NLI_UniversalClassifier_beta

nli_classifier

This model is a fine-tuned version of microsoft/deberta-v3-base on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.3715
• F1 Macro: 0.8526
• F1 Micro: 0.8656
• Accuracy Balanced: 0.8527
• Accuracy: 0.8656
• Precision Macro: 0.8526
• Recall Macro: 0.8527
• Precision Micro: 0.8656
• Recall Micro: 0.8656

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: 16
• eval_batch_size: 128
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 64
• 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_ratio: 0.06
• num_epochs: 2
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	F1 Macro	F1 Micro	Accuracy Balanced	Accuracy	Precision Macro	Recall Macro	Precision Micro	Recall Micro
0.2235	1.0	9145	0.3456	0.8493	0.8627	0.8488	0.8627	0.8497	0.8488	0.8627	0.8627
0.1575	1.9998	18288	0.3715	0.8526	0.8656	0.8527	0.8656	0.8526	0.8527	0.8656	0.8656

Framework versions

• Transformers 4.51.3
• Pytorch 2.5.1+cu124
• Datasets 3.5.0
• Tokenizers 0.21.0