led-real_risalah-version

This model is a fine-tuned version of silmiaulia/led-risalah-v2 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 2.7213
• Rouge1 Precision: 0.5502
• Rouge1 Recall: 0.1447
• Rouge1 Fmeasure: 0.229

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: 1
• eval_batch_size: 1
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 4
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 4
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1 Precision	Rouge1 Recall	Rouge1 Fmeasure
3.5241	1.0	6	3.0864	0.4855	0.1211	0.1932
2.7728	2.0	12	2.8611	0.4713	0.1193	0.1902
2.4564	3.0	18	2.7424	0.5664	0.1453	0.231
2.282	4.0	24	2.7213	0.5502	0.1447	0.229

Framework versions

• Transformers 4.41.2
• Pytorch 2.3.0+cu121
• Datasets 2.19.2
• Tokenizers 0.19.1