aafb442c8d0c5a9f5bb1c54a37bdf9d6

This model is a fine-tuned version of google-bert/bert-base-cased on the google/boolq dataset. It achieves the following results on the evaluation set:

• Loss: 0.9724
• Data Size: 1.0
• Epoch Runtime: 15.9380
• Accuracy: 0.7013
• F1 Macro: 0.6783
• Rouge1: 0.7019
• Rouge2: 0.0
• Rougel: 0.7007
• Rougelsum: 0.7010

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	Rouge1	Rouge2	Rougel	Rougelsum
No log	0	0	0.6696	0	1.9501	0.6213	0.3832	0.6213	0.0	0.6207	0.6210
No log	1	294	0.7166	0.0078	2.3474	0.4350	0.4063	0.4344	0.0	0.4357	0.4354
No log	2	588	0.6663	0.0156	2.3618	0.6204	0.3829	0.6204	0.0	0.6198	0.6203
No log	3	882	0.6672	0.0312	2.7839	0.6167	0.4350	0.6170	0.0	0.6160	0.6167
0.0272	4	1176	0.6601	0.0625	3.1876	0.625	0.4060	0.625	0.0	0.6244	0.625
0.0557	5	1470	0.6535	0.125	4.1122	0.6213	0.3832	0.6213	0.0	0.6207	0.6210
0.0922	6	1764	0.6255	0.25	5.8028	0.6419	0.5013	0.6419	0.0	0.6412	0.6419
0.581	7	2058	0.6102	0.5	9.2639	0.6703	0.6592	0.6703	0.0	0.6700	0.6706
0.5061	8.0	2352	0.5991	1.0	16.2840	0.6808	0.6653	0.6809	0.0	0.6801	0.6805
0.3405	9.0	2646	0.7041	1.0	15.8288	0.7093	0.6756	0.7093	0.0	0.7086	0.7093
0.1959	10.0	2940	0.8492	1.0	16.1055	0.7151	0.6852	0.7154	0.0	0.7148	0.7154
0.1583	11.0	3234	1.0167	1.0	15.9238	0.7142	0.6956	0.7138	0.0	0.7132	0.7142
0.1147	12.0	3528	0.9724	1.0	15.9380	0.7013	0.6783	0.7019	0.0	0.7007	0.7010

Framework versions

• Transformers 4.57.0
• Pytorch 2.8.0+cu128
• Datasets 4.3.0
• Tokenizers 0.22.1