train_copa_1753094177

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

• Loss: 0.1169
• Num Input Tokens Seen: 281856

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: 123
• 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
0.6395	0.5	45	0.6500	14016
0.7006	1.0	90	0.4485	28096
0.1984	1.5	135	0.1417	42144
0.1022	2.0	180	0.1372	56128
0.0941	2.5	225	0.1288	70272
0.0492	3.0	270	0.1268	84352
0.1109	3.5	315	0.1257	98464
0.0217	4.0	360	0.1230	112576
0.1727	4.5	405	0.1221	126624
0.0202	5.0	450	0.1202	140832
0.052	5.5	495	0.1207	154976
0.0191	6.0	540	0.1234	169056
0.1793	6.5	585	0.1185	183200
0.0722	7.0	630	0.1177	197344
0.0602	7.5	675	0.1186	211392
0.0334	8.0	720	0.1204	225536
0.0266	8.5	765	0.1173	239680
0.046	9.0	810	0.1169	253696
0.0042	9.5	855	0.1186	267840
0.0783	10.0	900	0.1192	281856

Framework versions

• PEFT 0.15.2
• Transformers 4.51.3
• Pytorch 2.7.1+cu126
• Datasets 3.6.0
• Tokenizers 0.21.1