led-large-16384-finetune-paperLedBASEAttention-BASEAttention3ep0.0

This model is a fine-tuned version of haryde/led-large-16384-finetune-paperLedBASEAttention on the None dataset. It achieves the following results on the evaluation set:

• Loss: 2.8780
• Rouge1: 32.8733
• Rouge2: 9.8113
• Rougel: 15.8093
• Rougelsum: 31.8695
• Gen Len: 1.0

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: 4
• eval_batch_size: 4
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 16
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• num_epochs: 3
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Gen Len
2.4328	0.9993	1128	2.8710	34.9112	9.4955	16.568	32.2485	1.0
2.4026	1.9993	2256	2.8637	36.5112	8.9613	18.2556	33.6714	1.0
2.3807	2.9993	3384	2.8780	32.8733	9.8113	15.8093	31.8695	1.0

Framework versions

• Transformers 4.48.3
• Pytorch 2.6.0+cu124
• Datasets 2.14.4
• Tokenizers 0.21.1