led-risalah_data_v9

This model is a fine-tuned version of silmi224/finetune-led-35000 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 1.6418
• Rouge1 Precision: 0.6262
• Rouge1 Recall: 0.3099
• Rouge1 Fmeasure: 0.4143

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: 10
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1 Precision	Rouge1 Recall	Rouge1 Fmeasure
1.2018	0.9714	17	1.4789	0.5782	0.2807	0.3761
1.0123	2.0	35	1.4305	0.5931	0.2892	0.3876
0.8845	2.9714	52	1.4693	0.6327	0.3088	0.4148
0.705	4.0	70	1.4903	0.6263	0.3052	0.4096
0.6323	4.9714	87	1.5086	0.6167	0.3052	0.4075
0.5926	6.0	105	1.5386	0.6238	0.3031	0.4072
0.5149	6.9714	122	1.5742	0.6308	0.3035	0.4096
0.4324	8.0	140	1.6112	0.6188	0.3083	0.411
0.3748	8.9714	157	1.6382	0.6262	0.3097	0.4138
0.4033	9.7143	170	1.6418	0.6262	0.3099	0.4143

Framework versions

• Transformers 4.41.2
• Pytorch 2.1.2
• Datasets 2.19.2
• Tokenizers 0.19.1