summary_train2

This model is a fine-tuned version of meta-llama/Meta-Llama-3-8B on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 2.6767

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: 4
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 16
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine
• num_epochs: 20
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
No log	1.0	17	0.9815
No log	2.0	34	0.9102
No log	3.0	51	0.8509
No log	4.0	68	0.8540
No log	5.0	85	0.8934
0.7594	6.0	102	0.9420
0.7594	7.0	119	1.0024
0.7594	8.0	136	1.1378
0.7594	9.0	153	1.2391
0.7594	10.0	170	1.3776
0.7594	11.0	187	1.5790
0.3001	12.0	204	1.8004
0.3001	13.0	221	2.0407
0.3001	14.0	238	2.2108
0.3001	15.0	255	2.3857
0.3001	16.0	272	2.5249
0.3001	17.0	289	2.6173
0.0996	18.0	306	2.6625
0.0996	19.0	323	2.6729
0.0996	20.0	340	2.6767

Framework versions

• PEFT 0.13.1
• Transformers 4.43.3
• Pytorch 2.4.0+cu121
• Datasets 3.0.1
• Tokenizers 0.19.1