51b72ab6465ea7dfc0a9db2012c52c86

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

• Loss: 0.8617
• Data Size: 1.0
• Epoch Runtime: 13.0217
• Accuracy: 0.8190
• F1 Macro: 0.7912
• Rouge1: 0.8196
• Rouge2: 0.0
• Rougel: 0.8196
• Rougelsum: 0.8190

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	0.6989	0	2.0856	0.4316	0.4285	0.4316	0.0	0.4316	0.4304
No log	1	114	0.6460	0.0078	3.6875	0.6651	0.3994	0.6657	0.0	0.6645	0.6651
No log	2	228	0.6776	0.0156	3.2836	0.6651	0.3994	0.6657	0.0	0.6645	0.6651
No log	3	342	0.6232	0.0312	3.8841	0.6651	0.3994	0.6657	0.0	0.6645	0.6651
0.0205	4	456	0.6074	0.0625	4.7494	0.6810	0.4681	0.6810	0.0	0.6810	0.6810
0.0205	5	570	0.5794	0.125	5.5135	0.7105	0.6015	0.7105	0.0	0.7105	0.7105
0.0205	6	684	0.5386	0.25	6.5594	0.7270	0.6177	0.7270	0.0	0.7270	0.7270
0.1308	7	798	0.4695	0.5	9.1107	0.8054	0.7666	0.8060	0.0	0.8054	0.8060
0.3834	8.0	912	0.3974	1.0	14.9734	0.8154	0.7790	0.8154	0.0	0.8154	0.8154
0.2399	9.0	1026	0.4312	1.0	14.1936	0.8190	0.7964	0.8196	0.0	0.8190	0.8196
0.1521	10.0	1140	0.6332	1.0	14.7622	0.8178	0.7849	0.8184	0.0	0.8172	0.8184
0.116	11.0	1254	0.7650	1.0	14.1674	0.8101	0.7678	0.8101	0.0	0.8096	0.8107
0.0802	12.0	1368	0.8617	1.0	13.0217	0.8190	0.7912	0.8196	0.0	0.8196	0.8190

Framework versions

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