1d8adfba175c877c0d15fc4289d7458e

This model is a fine-tuned version of meta-llama/Llama-3.2-3B on the contemmcm/amazon_reviews_2013 [cell-phone] dataset. It achieves the following results on the evaluation set:

• Loss: 4.7187
• Data Size: 1.0
• Epoch Runtime: 1051.5190
• Accuracy: 0.6387
• F1 Macro: 0.5707

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: 5e-05
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• distributed_type: multi-GPU
• num_devices: 4
• total_train_batch_size: 32
• total_eval_batch_size: 32
• optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: constant
• num_epochs: 50

Training results

Training Loss	Epoch	Step	Validation Loss	Data Size	Epoch Runtime	Accuracy	F1 Macro
No log	0	0	8.6758	0	67.8584	0.2345	0.1423
No log	1	1973	6.4235	0.0078	76.1140	0.2948	0.1473
0.2409	2	3946	6.2780	0.0156	85.0401	0.3853	0.1112
4.943	3	5919	4.6684	0.0312	103.7921	0.5096	0.3337
4.3992	4	7892	3.9270	0.0625	134.8878	0.5967	0.4505
3.6998	5	9865	3.5557	0.125	199.0520	0.6336	0.5083
3.6085	6	11838	3.6052	0.25	321.6011	0.6284	0.4421
3.5503	7	13811	3.2257	0.5	568.6083	0.6653	0.5589
3.2269	8.0	15784	3.2006	1.0	1056.6024	0.6601	0.5812
2.6516	9.0	17757	3.4675	1.0	1049.3552	0.6547	0.5335
2.296	10.0	19730	3.5568	1.0	1053.2807	0.6335	0.5846
1.7596	11.0	21703	4.1567	1.0	1053.0386	0.6430	0.5651
1.4382	12.0	23676	4.7187	1.0	1051.5190	0.6387	0.5707

Framework versions

• Transformers 4.57.0
• Pytorch 2.8.0+cu128
• Datasets 4.2.0
• Tokenizers 0.22.1