CuATR-distilbert-LoRA

This model is a fine-tuned version of distilbert-base-uncased on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.6945
• Accuracy: 0.5652
• F1: 0.7222

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: 16
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 64
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 14
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1
0.7007	0.67	1	0.6957	0.5652	0.7222
0.6984	2.0	3	0.6954	0.5652	0.7222
0.7075	2.67	4	0.6952	0.5652	0.7222
0.7	4.0	6	0.6951	0.5652	0.7222
0.6962	4.67	7	0.6950	0.5652	0.7222
0.7003	6.0	9	0.6948	0.5652	0.7222
0.6952	6.67	10	0.6947	0.5652	0.7222
0.7027	8.0	12	0.6946	0.5652	0.7222
0.6995	8.67	13	0.6946	0.5652	0.7222
0.6919	9.33	14	0.6945	0.5652	0.7222

Framework versions

• Transformers 4.35.2
• Pytorch 2.1.1+cu118
• Datasets 2.15.0
• Tokenizers 0.15.0