train_cola_456_1760637815

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

• Loss: 0.5826
• Num Input Tokens Seen: 6508096

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
0.2669	2.0	3420	0.2578	650880
0.2632	4.0	6840	0.2559	1301472
0.2325	6.0	10260	0.2594	1953024
0.3018	8.0	13680	0.2786	2602848
0.2229	10.0	17100	0.2923	3253216
0.294	12.0	20520	0.2974	3904704
0.118	14.0	23940	0.3779	4555936
0.077	16.0	27360	0.4856	5206016
0.0223	18.0	30780	0.5542	5857472
0.0609	20.0	34200	0.5826	6508096

Framework versions

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