d891af3fa7494f41c32c525c3589a4ac

This model is a fine-tuned version of google-bert/bert-large-uncased on the contemmcm/amazon_reviews_2013 [cell-phone] dataset. It achieves the following results on the evaluation set:

• Loss: 0.8539
• Data Size: 1.0
• Epoch Runtime: 228.8343
• Accuracy: 0.6922
• F1 Macro: 0.6146

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	1.6278	0	12.5732	0.2255	0.0736
No log	1	1973	1.3113	0.0078	15.5834	0.5082	0.2609
0.0281	2	3946	1.0342	0.0156	16.2502	0.5834	0.3934
0.9584	3	5919	1.0014	0.0312	20.8304	0.5884	0.4171
0.8509	4	7892	0.8080	0.0625	29.0264	0.6606	0.5723
0.8182	5	9865	0.8195	0.125	40.7370	0.6594	0.5545
0.785	6	11838	0.7900	0.25	67.2397	0.6710	0.5937
0.8295	7	13811	0.8114	0.5	120.3256	0.6784	0.5776
0.7436	8.0	15784	0.7499	1.0	228.2865	0.6834	0.6180
0.6975	9.0	17757	0.7628	1.0	225.8392	0.6859	0.6168
0.5624	10.0	19730	0.7665	1.0	226.6270	0.6756	0.6288
0.5489	11.0	21703	0.8420	1.0	228.3788	0.6864	0.6119
0.457	12.0	23676	0.8539	1.0	228.8343	0.6922	0.6146

Framework versions

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