train_siqa_123_1760637714

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

• Loss: 1.1770
• Num Input Tokens Seen: 53571360

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: 1e-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
0.3864	2.0	13364	0.4281	5355648
0.1761	4.0	26728	0.2439	10712224
0.0805	6.0	40092	0.2658	16069856
0.0835	8.0	53456	0.3968	21425728
0.1285	10.0	66820	0.5146	26781984
0.0009	12.0	80184	0.6194	32139680
0.0	14.0	93548	0.8041	37497632
0.1594	16.0	106912	0.9499	42854976
0.0	18.0	120276	1.0988	48213248
0.0	20.0	133640	1.1770	53571360

Framework versions

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