c3a05faacbdd7f13361e1d2c28d3df32

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

• Loss: 0.6433
• Data Size: 1.0
• Epoch Runtime: 26.1368
• Accuracy: 0.6651
• F1 Macro: 0.3994
• Rouge1: 0.6657
• Rouge2: 0.0
• Rougel: 0.6645
• Rougelsum: 0.6651

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	0.6722	0	3.2975	0.6651	0.3994	0.6657	0.0	0.6645	0.6651
No log	1	114	0.6769	0.0078	3.7979	0.6651	0.3994	0.6657	0.0	0.6645	0.6651
No log	2	228	0.6578	0.0156	4.8465	0.6651	0.3994	0.6657	0.0	0.6645	0.6651
No log	3	342	0.6445	0.0312	6.2523	0.6651	0.3994	0.6657	0.0	0.6645	0.6651
0.0212	4	456	0.6337	0.0625	7.4135	0.6651	0.3994	0.6657	0.0	0.6645	0.6651
0.0212	5	570	0.5898	0.125	9.1441	0.6651	0.3994	0.6657	0.0	0.6645	0.6651
0.0212	6	684	0.4361	0.25	11.9615	0.8196	0.7898	0.8196	0.0	0.8196	0.8196
0.1356	7	798	0.4029	0.5	17.4409	0.8302	0.8048	0.8308	0.0	0.8296	0.8308
0.4019	8.0	912	0.4426	1.0	28.6050	0.8184	0.7722	0.8184	0.0	0.8184	0.8190
0.268	9.0	1026	0.3874	1.0	26.7759	0.8426	0.8323	0.8426	0.0	0.8426	0.8432
0.2538	10.0	1140	0.4039	1.0	26.2096	0.8237	0.8145	0.8243	0.0	0.8243	0.8237
0.5484	11.0	1254	0.6329	1.0	26.4826	0.6851	0.4636	0.6857	0.0	0.6851	0.6851
0.6496	12.0	1368	0.6421	1.0	26.1090	0.6651	0.3994	0.6657	0.0	0.6645	0.6651
0.6459	13.0	1482	0.6433	1.0	26.1368	0.6651	0.3994	0.6657	0.0	0.6645	0.6651

Framework versions

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