10ee4872362eb2c341e638c0da24616a

This model is a fine-tuned version of FacebookAI/xlm-roberta-large-finetuned-conll02-dutch on the dim/tldr_news dataset. It achieves the following results on the evaluation set:

• Loss: 0.8541
• Data Size: 1.0
• Epoch Runtime: 32.6044
• Accuracy: 0.7472
• F1 Macro: 0.7862
• Rouge1: 0.7472
• Rouge2: 0.0
• Rougel: 0.7479
• Rougelsum: 0.7479

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	1.7142	0	2.7923	0.2145	0.0782	0.2138	0.0	0.2145	0.2145
No log	1	178	1.5677	0.0078	3.2286	0.2401	0.0774	0.2401	0.0	0.2408	0.2393
No log	2	356	1.4449	0.0156	3.6535	0.3679	0.2026	0.3679	0.0	0.3679	0.3679
No log	3	534	1.1423	0.0312	5.0216	0.5270	0.3708	0.5277	0.0	0.5270	0.5277
No log	4	712	0.9032	0.0625	6.6524	0.6854	0.5321	0.6857	0.0	0.6861	0.6854
No log	5	890	0.9358	0.125	8.9451	0.6825	0.5163	0.6832	0.0	0.6825	0.6825
0.0673	6	1068	0.8347	0.25	12.5451	0.7088	0.5679	0.7088	0.0	0.7092	0.7088
0.6965	7	1246	0.7293	0.5	19.3701	0.7372	0.5662	0.7379	0.0	0.7379	0.7365
0.566	8.0	1424	0.6066	1.0	34.2283	0.7607	0.7639	0.7607	0.0	0.7621	0.7607
0.5462	9.0	1602	0.6888	1.0	32.6539	0.7550	0.7886	0.7553	0.0	0.7557	0.7550
0.41	10.0	1780	0.6840	1.0	32.7643	0.7599	0.7845	0.7607	0.0	0.7607	0.7607
0.3802	11.0	1958	0.7838	1.0	32.8990	0.7401	0.7741	0.7401	0.0	0.7408	0.7401
0.3008	12.0	2136	0.8541	1.0	32.6044	0.7472	0.7862	0.7472	0.0	0.7479	0.7479

Framework versions

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