train_svamp_123_1760637659

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

• Loss: 0.0751
• Num Input Tokens Seen: 1433472

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: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Input Tokens Seen
0.0882	1.0	158	0.0970	71776
0.077	2.0	316	0.0751	143552
0.012	3.0	474	0.0920	215312
0.0436	4.0	632	0.0957	286736
0.019	5.0	790	0.1062	358416
0.014	6.0	948	0.1438	430224
0.0002	7.0	1106	0.1349	501728
0.0001	8.0	1264	0.1548	573328
0.0006	9.0	1422	0.1386	645088
0.0	10.0	1580	0.1408	716976
0.0	11.0	1738	0.1438	788752
0.0	12.0	1896	0.1447	860352
0.0	13.0	2054	0.1473	931872
0.0	14.0	2212	0.1472	1003600
0.0	15.0	2370	0.1478	1075312
0.0	16.0	2528	0.1509	1146848
0.0	17.0	2686	0.1496	1218416
0.0	18.0	2844	0.1495	1290080
0.0	19.0	3002	0.1499	1361712
0.0	20.0	3160	0.1492	1433472

Framework versions

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