8bd0d648d782cdfaadb766c4dd3c57f4

This model is a fine-tuned version of distilbert/distilbert-base-multilingual-cased on the dim/tldr_news dataset. It achieves the following results on the evaluation set:

• Loss: 1.3080
• Data Size: 1.0
• Epoch Runtime: 7.4784
• Accuracy: 0.7571
• F1 Macro: 0.7781
• Rouge1: 0.7578
• Rouge2: 0.0
• Rougel: 0.7578
• Rougelsum: 0.7578

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	1.6181	0	0.9829	0.2429	0.0942	0.2436	0.0	0.2436	0.2429
No log	1	178	1.5573	0.0078	2.2929	0.2401	0.0774	0.2401	0.0	0.2408	0.2393
No log	2	356	1.4640	0.0156	1.4113	0.2401	0.0774	0.2401	0.0	0.2408	0.2393
No log	3	534	1.2233	0.0312	1.8476	0.5064	0.3366	0.5071	0.0	0.5060	0.5064
No log	4	712	0.8985	0.0625	2.3746	0.6669	0.4762	0.6683	0.0	0.6676	0.6669
No log	5	890	0.7910	0.125	2.8544	0.6974	0.5193	0.6982	0.0	0.6974	0.6974
0.0611	6	1068	0.7564	0.25	3.1930	0.6974	0.5471	0.6982	0.0	0.6982	0.6974
0.5937	7	1246	0.6769	0.5	4.5793	0.7415	0.7151	0.7415	0.0	0.7422	0.7415
0.4952	8.0	1424	0.6170	1.0	7.8104	0.7507	0.7394	0.7514	0.0	0.7507	0.7507
0.3554	9.0	1602	0.7022	1.0	7.5035	0.7621	0.7792	0.7621	0.0	0.7631	0.7621
0.1632	10.0	1780	0.8514	1.0	7.5236	0.7564	0.7774	0.7571	0.0	0.7564	0.7560
0.1036	11.0	1958	1.0957	1.0	7.5637	0.7521	0.7751	0.7528	0.0	0.7528	0.7521
0.0871	12.0	2136	1.3080	1.0	7.4784	0.7571	0.7781	0.7578	0.0	0.7578	0.7578

Framework versions

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