e99068e187339f9d0b0b2d345a20ed00

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

• Loss: 0.6931
• Data Size: 1.0
• Epoch Runtime: 570.9415
• Accuracy: 0.5057
• F1 Macro: 0.3359
• Rouge1: 0.5053
• Rouge2: 0.0
• Rougel: 0.5057
• Rougelsum: 0.5056

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	0.6961	0	8.9883	0.5057	0.3359	0.5053	0.0	0.5057	0.5056
No log	1	3273	0.6932	0.0078	14.0998	0.5	0.4732	0.5	0.0	0.4998	0.5002
0.0115	2	6546	0.6963	0.0156	18.9905	0.5057	0.3359	0.5053	0.0	0.5057	0.5056
0.6993	3	9819	0.6961	0.0312	27.7492	0.4943	0.3308	0.4947	0.0	0.4943	0.4944
0.7055	4	13092	0.6969	0.0625	45.5816	0.5057	0.3359	0.5053	0.0	0.5057	0.5056
0.6997	5	16365	0.6930	0.125	81.2118	0.4943	0.3308	0.4947	0.0	0.4943	0.4944
0.7006	6	19638	0.6956	0.25	150.3338	0.5057	0.3359	0.5053	0.0	0.5057	0.5056
0.6986	7	22911	0.6938	0.5	290.1053	0.4943	0.3308	0.4947	0.0	0.4943	0.4944
0.6957	8.0	26184	0.6939	1.0	565.6960	0.4943	0.3308	0.4947	0.0	0.4943	0.4944
0.6986	9.0	29457	0.6931	1.0	570.9415	0.5057	0.3359	0.5053	0.0	0.5057	0.5056

Framework versions

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