985f0ddda38bd171d8d049c44911e08b

This model is a fine-tuned version of distilbert/distilgpt2 on the fancyzhx/dbpedia_14 dataset. It achieves the following results on the evaluation set:

• Loss: 0.0559
• Data Size: 1.0
• Epoch Runtime: 737.9425
• Accuracy: 0.9907
• F1 Macro: 0.9907
• Rouge1: 0.9907
• Rouge2: 0.0
• Rougel: 0.9907
• Rougelsum: 0.9907

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	5.4449	0	34.8319	0.0731	0.0129	0.0730	0.0	0.0731	0.0731
0.2379	1	17500	0.1123	0.0078	41.9065	0.9708	0.9706	0.9708	0.0	0.9708	0.9708
0.0768	2	35000	0.0843	0.0156	45.7037	0.9816	0.9816	0.9816	0.0	0.9816	0.9816
0.0465	3	52500	0.0770	0.0312	56.3884	0.9833	0.9833	0.9834	0.0	0.9833	0.9833
0.0693	4	70000	0.0579	0.0625	80.4065	0.9864	0.9864	0.9864	0.0	0.9864	0.9864
0.0417	5	87500	0.0571	0.125	124.0997	0.9880	0.9880	0.9880	0.0	0.9880	0.9880
0.0547	6	105000	0.0471	0.25	206.0458	0.9889	0.9889	0.9889	0.0	0.9889	0.9889
0.0003	7	122500	0.0401	0.5	371.8162	0.9904	0.9904	0.9904	0.0	0.9904	0.9904
0.0265	8.0	140000	0.0373	1.0	738.3664	0.9913	0.9913	0.9913	0.0	0.9913	0.9913
0.0143	9.0	157500	0.0409	1.0	737.4335	0.9908	0.9908	0.9908	0.0	0.9908	0.9908
0.0183	10.0	175000	0.0442	1.0	739.0930	0.9913	0.9913	0.9913	0.0	0.9912	0.9913
0.0144	11.0	192500	0.0525	1.0	740.0167	0.9913	0.9913	0.9913	0.0	0.9913	0.9913
0.0076	12.0	210000	0.0559	1.0	737.9425	0.9907	0.9907	0.9907	0.0	0.9907	0.9907

Framework versions

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