d0a2530e1d5cbdda4bbeb62fc3c90415

This model is a fine-tuned version of google-bert/bert-large-cased-whole-word-masking on the fancyzhx/dbpedia_14 dataset. It achieves the following results on the evaluation set:

• Loss: 0.0844
• Data Size: 0.25
• Epoch Runtime: 437.3712
• Accuracy: 0.9860
• F1 Macro: 0.9860
• Rouge1: 0.9860
• Rouge2: 0.0
• Rougel: 0.9860
• Rougelsum: 0.9860

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	2.7507	0	53.4048	0.0702	0.0286	0.0702	0.0	0.0702	0.0702
0.1093	1	17500	0.0732	0.0078	66.4411	0.9844	0.9843	0.9844	0.0	0.9844	0.9844
0.0603	2	35000	0.0599	0.0156	80.9481	0.9884	0.9884	0.9884	0.0	0.9884	0.9884
0.0643	3	52500	0.0937	0.0312	102.8833	0.9831	0.9831	0.9831	0.0	0.9830	0.9831
0.0652	4	70000	0.0751	0.0625	151.7207	0.9861	0.9862	0.9862	0.0	0.9861	0.9861
0.0754	5	87500	0.0637	0.125	247.3297	0.9876	0.9876	0.9876	0.0	0.9876	0.9876
0.109	6	105000	0.0844	0.25	437.3712	0.9860	0.9860	0.9860	0.0	0.9860	0.9860

Framework versions

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