train_mrpc_456_1760637792

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

• Loss: 0.2346
• Num Input Tokens Seen: 6773216

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: 456
• 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: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Input Tokens Seen
0.2321	1.0	826	0.2240	338864
0.2171	2.0	1652	0.1761	676984
0.1427	3.0	2478	0.1641	1016176
0.2379	4.0	3304	0.1589	1354632
0.1037	5.0	4130	0.1515	1692816
0.1423	6.0	4956	0.1513	2031320
0.0681	7.0	5782	0.1670	2369768
0.108	8.0	6608	0.1435	2708688
0.0179	9.0	7434	0.1553	3047376
0.0706	10.0	8260	0.1431	3386408
0.0573	11.0	9086	0.1348	3724296
0.1185	12.0	9912	0.1581	4063352
0.0091	13.0	10738	0.1886	4402032
0.0026	14.0	11564	0.2011	4740464
0.0182	15.0	12390	0.2893	5079384
0.0029	16.0	13216	0.3326	5418192
0.0006	17.0	14042	0.3731	5757208
0.0005	18.0	14868	0.3943	6095648
0.0208	19.0	15694	0.4036	6434448
0.0008	20.0	16520	0.4064	6773216

Framework versions

• PEFT 0.17.1
• Transformers 4.51.3
• Pytorch 2.9.0+cu128
• Datasets 4.0.0
• Tokenizers 0.21.4