distilbert-base-uncased-lora-text-classification

This model is a fine-tuned version of distilbert-base-uncased on the None dataset. It achieves the following results on the evaluation set:

• Loss: 1.6464
• Accuracy: {'accuracy': 0.84}

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: 0.001
• train_batch_size: 4
• eval_batch_size: 4
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	1.0	250	0.5728	{'accuracy': 0.824}
0.4384	2.0	500	0.6861	{'accuracy': 0.84}
0.4384	3.0	750	1.1608	{'accuracy': 0.812}
0.1949	4.0	1000	1.0198	{'accuracy': 0.826}
0.1949	5.0	1250	1.1314	{'accuracy': 0.838}
0.0612	6.0	1500	1.3810	{'accuracy': 0.844}
0.0612	7.0	1750	1.6426	{'accuracy': 0.832}
0.0164	8.0	2000	1.6141	{'accuracy': 0.844}
0.0164	9.0	2250	1.6768	{'accuracy': 0.842}
0.0075	10.0	2500	1.6464	{'accuracy': 0.84}

Framework versions

• PEFT 0.12.0
• Transformers 4.42.4
• Pytorch 2.3.1
• Datasets 2.20.0
• Tokenizers 0.19.1