1f133c773edd97e29a2555290554c261

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

• Loss: 0.7872
• Data Size: 1.0
• Epoch Runtime: 14.2633
• Accuracy: 0.8078
• F1 Macro: 0.7655
• Rouge1: 0.8078
• Rouge2: 0.0
• Rougel: 0.8078
• Rougelsum: 0.8081

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	0.6632	0	1.8926	0.6621	0.4017	0.6627	0.0	0.6621	0.6621
No log	1	114	0.6403	0.0078	2.8475	0.6598	0.4104	0.6604	0.0	0.6592	0.6592
No log	2	228	0.6440	0.0156	2.4447	0.6651	0.3994	0.6657	0.0	0.6645	0.6651
No log	3	342	0.6271	0.0312	2.9577	0.6651	0.3994	0.6657	0.0	0.6645	0.6651
0.0204	4	456	0.6055	0.0625	3.8269	0.6881	0.5313	0.6884	0.0	0.6881	0.6875
0.0204	5	570	0.6103	0.125	4.5648	0.6798	0.4580	0.6798	0.0	0.6798	0.6798
0.0204	6	684	0.4979	0.25	6.2226	0.7429	0.7041	0.7429	0.0	0.7429	0.7435
0.1357	7	798	0.5088	0.5	8.7492	0.7241	0.7207	0.7235	0.0	0.7246	0.7241
0.3768	8.0	912	0.5057	1.0	15.1075	0.7423	0.6234	0.7429	0.0	0.7423	0.7423
0.2825	9.0	1026	0.4990	1.0	15.6576	0.8019	0.7445	0.8019	0.0	0.8019	0.8019
0.352	10.0	1140	0.4733	1.0	14.6845	0.8125	0.7768	0.8125	0.0	0.8125	0.8131
0.3582	11.0	1254	0.5672	1.0	14.2787	0.8219	0.8020	0.8219	0.0	0.8219	0.8219
0.3769	12.0	1368	0.8509	1.0	14.7253	0.8202	0.8009	0.8202	0.0	0.8208	0.8196
0.1844	13.0	1482	0.6314	1.0	14.6650	0.7860	0.7538	0.7860	0.0	0.7860	0.7854
0.198	14.0	1596	0.7872	1.0	14.2633	0.8078	0.7655	0.8078	0.0	0.8078	0.8081

Framework versions

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