628f7f0a34a6d4e3a69724676ecbc9b1

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

• Loss: 0.4231
• Data Size: 1.0
• Epoch Runtime: 31.3019
• Accuracy: 0.9062
• F1 Macro: 0.9112

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
No log	0	0	1.8950	0	1.6357	0.1792	0.0506
No log	1	170	1.6924	0.0078	2.2008	0.2875	0.1010
No log	2	340	1.7226	0.0156	2.7158	0.1333	0.0392
No log	3	510	1.6410	0.0312	4.0701	0.1812	0.0564
No log	4	680	1.1819	0.0625	5.6504	0.5625	0.4145
0.0826	5	850	0.4079	0.125	7.8871	0.9021	0.7545
0.0826	6	1020	0.4596	0.25	11.9577	0.8938	0.7475
0.507	7	1190	0.5043	0.5	18.5672	0.875	0.8660
0.4148	8.0	1360	0.4105	1.0	31.7160	0.8812	0.7429
0.3695	9.0	1530	0.4231	1.0	31.3019	0.9062	0.9112

Framework versions

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