5ab5445f9501080690f22739d99457a2

This model is a fine-tuned version of google-bert/bert-large-cased on the nyu-mll/glue [mrpc] dataset. It achieves the following results on the evaluation set:

• Loss: 0.5478
• Data Size: 1.0
• Epoch Runtime: 14.7532
• Accuracy: 0.7966
• F1 Macro: 0.7450
• Rouge1: 0.7966
• Rouge2: 0.0
• Rougel: 0.7972
• Rougelsum: 0.7972

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	0.7767	0	1.8801	0.3379	0.2565	0.3373	0.0	0.3384	0.3379
No log	1	114	0.6388	0.0078	2.3708	0.6604	0.4258	0.6610	0.0	0.6598	0.6604
No log	2	228	0.6437	0.0156	2.6195	0.6651	0.3994	0.6657	0.0	0.6645	0.6651
No log	3	342	0.6426	0.0312	3.2657	0.6645	0.3992	0.6651	0.0	0.6645	0.6645
0.0214	4	456	0.6652	0.0625	4.2861	0.6651	0.3994	0.6657	0.0	0.6645	0.6651
0.0214	5	570	0.6356	0.125	5.1791	0.6651	0.3994	0.6657	0.0	0.6645	0.6651
0.0214	6	684	0.5730	0.25	6.3667	0.6739	0.4409	0.6739	0.0	0.6739	0.6739
0.1519	7	798	0.5876	0.5	8.9274	0.6728	0.6716	0.6728	0.0	0.6731	0.6728
0.4537	8.0	912	0.4532	1.0	15.0116	0.7948	0.7422	0.7954	0.0	0.7948	0.7948
0.4538	9.0	1026	0.5870	1.0	14.6384	0.7459	0.6525	0.7465	0.0	0.7459	0.7465
0.3925	10.0	1140	0.5264	1.0	14.0695	0.7807	0.7637	0.7807	0.0	0.7807	0.7812
0.4673	11.0	1254	0.4680	1.0	14.2278	0.8172	0.7877	0.8178	0.0	0.8178	0.8178
0.2974	12.0	1368	0.5478	1.0	14.7532	0.7966	0.7450	0.7966	0.0	0.7972	0.7972

Framework versions

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