train_rte_456_1760637784

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

• Loss: 1.1027
• Num Input Tokens Seen: 6215968

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: 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
0.1312	2.0	996	0.1648	622720
0.1786	4.0	1992	0.1621	1242912
0.1139	6.0	2988	0.1923	1862848
0.0842	8.0	3984	0.1915	2487712
0.0943	10.0	4980	0.2987	3108288
0.1454	12.0	5976	0.5770	3732864
0.0009	14.0	6972	0.8435	4355328
0.0001	16.0	7968	1.0402	4976032
0.0	18.0	8964	1.0930	5594752
0.0001	20.0	9960	1.1027	6215968

Framework versions

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