Modelpublic

ppak10/defect-classification-llama-baseline-15-epochs

sourceHugging Facellama3.2updated 2y agoView on Hugging Face

0likes

Model Card

defect-classification-llama-baseline-15-epochs

This model is a fine-tuned version of meta-llama/Llama-3.2-1B on an unknown dataset. It achieves the following results on the evaluation set:

—Loss: 0.1984
—Accuracy: 0.9311

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: 2e-05
—trainbatchsize: 512
—evalbatchsize: 512
—seed: 42
—optimizer: Use adamwtorch with betas=(0.9,0.999) and epsilon=1e-08 and optimizerargs=No additional optimizer arguments
—lrschedulertype: linear
—num_epochs: 15

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.8391	1.0	1062	0.8638	0.8010
0.5603	2.0	2124	0.5259	0.8492
0.4341	3.0	3186	0.4449	0.8659
0.3805	4.0	4248	0.4048	0.8587
0.3143	5.0	5310	0.3020	0.8974
0.2891	6.0	6372	0.2881	0.9011
0.2728	7.0	7434	0.2790	0.9036
0.305	8.0	8496	0.2568	0.9114
0.2382	9.0	9558	0.2377	0.9179
0.2258	10.0	10620	0.2202	0.9229
0.2202	11.0	11682	0.2180	0.9248
0.2153	12.0	12744	0.2249	0.9234
0.2026	13.0	13806	0.2039	0.9294
0.1972	14.0	14868	0.2009	0.9293
0.1987	15.0	15930	0.1984	0.9311

Framework versions

—Transformers 4.47.0
—Pytorch 2.5.1+cu124
—Datasets 3.2.0
—Tokenizers 0.21.0