train_cola_1757340262

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

• Loss: 0.1526
• Num Input Tokens Seen: 3663392

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: 5e-05
• train_batch_size: 4
• eval_batch_size: 4
• seed: 101112
• optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: cosine
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 10.0

Training results

Training Loss	Epoch	Step	Validation Loss	Input Tokens Seen
0.2238	0.5	962	0.1899	183040
0.3845	1.0	1924	0.1614	366136
0.3468	1.5	2886	0.1578	548856
0.0904	2.0	3848	0.1526	732880
0.2354	2.5	4810	0.1603	916368
0.0322	3.0	5772	0.1643	1099816
0.094	3.5	6734	0.1670	1283208
0.1749	4.0	7696	0.1753	1465464
0.0737	4.5	8658	0.1710	1648696
0.1082	5.0	9620	0.1681	1831728
0.1139	5.5	10582	0.1820	2014288
0.1319	6.0	11544	0.1667	2198176
0.0087	6.5	12506	0.1792	2382016
0.1629	7.0	13468	0.1827	2564208
0.1174	7.5	14430	0.1833	2746960
0.0416	8.0	15392	0.1861	2930240
0.2142	8.5	16354	0.1899	3113696
0.1691	9.0	17316	0.1846	3297136
0.1169	9.5	18278	0.1874	3480592
0.0104	10.0	19240	0.1869	3663392

Framework versions

• PEFT 0.15.2
• Transformers 4.51.3
• Pytorch 2.8.0+cu128
• Datasets 3.6.0
• Tokenizers 0.21.1