Llama3_8B_Task2_semantic_pred

This model is a fine-tuned version of meta-llama/Meta-Llama-3-8B on the None dataset. It achieves the following results on the evaluation set:

• Loss: 1.2767
• Accuracy: 0.6493
• Precision: 0.6493
• Recall: 0.6493
• F1 score: 0.6493

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: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 5

Training results

Training Loss	Epoch	Step	Accuracy	F1 score	Precision	Recall	Validation Loss
0.49	0.5208	200	0.5750	0.5750	0.5750	0.5750	0.9015
0.439	1.0417	400	0.5541	0.5541	0.5541	0.5541	1.2361
0.2744	1.5625	600	0.7744	0.7744	0.7744	0.7744	0.4804
0.2621	2.0833	800	0.5658	0.5658	0.5658	0.5658	1.2460
0.1921	2.6042	1000	0.6102	0.6102	0.6102	0.6102	1.0217
0.1602	3.125	1200	0.5880	0.5880	0.5880	0.5880	1.3196
0.1736	3.6458	1400	0.5684	0.5684	0.5684	0.5684	1.7235
0.1628	4.1667	1600	0.6780	0.6780	0.6780	0.6780	1.0542
0.1204	4.6875	1800	1.2767	0.6493	0.6493	0.6493	0.6493

Framework versions

• PEFT 0.11.1
• Transformers 4.44.2
• Pytorch 2.3.0+cu121
• Datasets 2.19.1
• Tokenizers 0.19.1