e94461670c5587c7c99dfd6190588bf6

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

• Loss: 1.0316
• Data Size: 1.0
• Epoch Runtime: 57.8210
• Accuracy: 0.6729
• F1 Macro: 0.6128

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	1.6343	0	4.7721	0.2068	0.1097
No log	1	1973	1.4304	0.0078	5.5269	0.4582	0.2730
0.0289	2	3946	1.0808	0.0156	5.4952	0.5523	0.3702
1.0124	3	5919	1.0449	0.0312	6.5121	0.5700	0.3720
0.8858	4	7892	0.8739	0.0625	8.1944	0.6305	0.4997
0.8288	5	9865	0.8130	0.125	11.2886	0.6582	0.5522
0.7711	6	11838	0.7932	0.25	18.3471	0.6694	0.5777
0.7857	7	13811	0.7642	0.5	31.1192	0.6800	0.6197
0.7052	8.0	15784	0.7654	1.0	58.4533	0.6848	0.6232
0.5752	9.0	17757	0.7964	1.0	57.0434	0.6885	0.6254
0.4501	10.0	19730	0.9360	1.0	57.2006	0.6563	0.6136
0.3683	11.0	21703	1.0316	1.0	57.8210	0.6729	0.6128

Framework versions

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