train_hellaswag_456_1760637853

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

• Loss: 0.5258
• Num Input Tokens Seen: 194107168

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
0.4696	2.0	15962	0.4627	19404608
0.4645	4.0	31924	0.4628	38820832
0.453	6.0	47886	0.4631	58233056
0.4619	8.0	63848	0.4628	77656576
0.4624	10.0	79810	0.4628	97057728
0.4572	12.0	95772	0.4654	116467904
0.4518	14.0	111734	0.4826	135879680
0.4148	16.0	127696	0.4988	155294048
0.5048	18.0	143658	0.5134	174700864
0.4003	20.0	159620	0.5258	194107168

Framework versions

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