ee275d15bf827097119aec68c2ae1622

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

• Loss: 0.6953
• Data Size: 0.125
• Epoch Runtime: 82.7248
• Accuracy: 0.4943
• F1 Macro: 0.3308
• Rouge1: 0.4947
• Rouge2: 0.0
• Rougel: 0.4943
• Rougelsum: 0.4944

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	0.6962	0	8.9319	0.4732	0.4450	0.4730	0.0	0.4732	0.4730
No log	1	3273	0.6100	0.0078	13.4561	0.7353	0.7337	0.7353	0.0	0.7353	0.7351
0.0112	2	6546	0.7399	0.0156	18.8455	0.5099	0.3462	0.5094	0.0	0.5097	0.5097
0.7339	3	9819	0.7223	0.0312	28.7082	0.5057	0.3359	0.5053	0.0	0.5057	0.5056
0.6992	4	13092	0.6972	0.0625	46.3079	0.4943	0.3308	0.4947	0.0	0.4943	0.4944
0.6988	5	16365	0.6953	0.125	82.7248	0.4943	0.3308	0.4947	0.0	0.4943	0.4944

Framework versions

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