train_copa_123_1760637643

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: 1.4203
• Num Input Tokens Seen: 500896

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
0.2394	2.0	160	0.2486	50112
0.2201	4.0	320	0.2382	100160
0.2344	6.0	480	0.2459	150272
0.2218	8.0	640	0.2801	200576
0.1573	10.0	800	0.3157	250496
0.1662	12.0	960	0.6243	300416
0.029	14.0	1120	1.1238	350464
0.0004	16.0	1280	1.3742	400608
0.0018	18.0	1440	1.3974	450688
0.0004	20.0	1600	1.4203	500896

Framework versions

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