train_record_123_1762712912

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

• Loss: 0.7540
• Num Input Tokens Seen: 825656008

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
0.2951	2.0	55542	0.3415	82555432
0.1998	4.0	111084	0.3146	165106712
0.2415	6.0	166626	0.3147	247670040
0.1763	8.0	222168	0.3450	330225160
0.1449	10.0	277710	0.3864	412808752
0.1226	12.0	333252	0.4621	495376336
0.2367	14.0	388794	0.5068	577949296
0.1024	16.0	444336	0.6149	660502512
0.1127	18.0	499878	0.7038	743065464
0.0973	20.0	555420	0.7540	825656008

Framework versions

• PEFT 0.15.2
• Transformers 4.51.3
• Pytorch 2.8.0+cu128
• Datasets 3.6.0
• Tokenizers 0.21.1