c2899febc5be23c20d7d5d4445fb5add

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

• Loss: 0.9409
• Data Size: 1.0
• Epoch Runtime: 111.9639
• Accuracy: 0.6640
• F1 Macro: 0.5850

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	1.6312	0	7.7141	0.0939	0.0344
No log	1	1973	1.4666	0.0078	8.4488	0.3946	0.1612
0.0315	2	3946	1.2491	0.0156	9.1027	0.4832	0.2362
1.1292	3	5919	1.0509	0.0312	10.8002	0.5721	0.3923
1.0205	4	7892	0.9421	0.0625	14.1503	0.6134	0.5011
0.9263	5	9865	0.8755	0.125	20.7728	0.6353	0.5282
0.8422	6	11838	0.8371	0.25	34.9402	0.6547	0.5380
0.8435	7	13811	0.7905	0.5	58.6505	0.6747	0.5833
0.7869	8.0	15784	0.7987	1.0	110.9934	0.6635	0.5936
0.6664	9.0	17757	0.7959	1.0	109.7067	0.6801	0.6100
0.5869	10.0	19730	0.8469	1.0	109.7053	0.6426	0.5985
0.5624	11.0	21703	0.9409	1.0	111.9639	0.6640	0.5850

Framework versions

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