mt5-base-finetuned-xsum

This model is a fine-tuned version of google/mt5-base on the None dataset. It achieves the following results on the evaluation set:

• Loss: nan
• Rouge1: 1.5134
• Rouge2: 0.2001
• Rougel: 1.4917
• Rougelsum: 1.4788
• Gen Len: 8.6992

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: 1e-06
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• num_epochs: 50
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Gen Len
0.0	1.0	559	nan	1.5134	0.2001	1.4917	1.4788	8.6992
0.0	2.0	1118	nan	1.5134	0.2001	1.4917	1.4788	8.6992
0.0	3.0	1677	nan	1.5134	0.2001	1.4917	1.4788	8.6992
0.0	4.0	2236	nan	1.5134	0.2001	1.4917	1.4788	8.6992
0.0	5.0	2795	nan	1.5134	0.2001	1.4917	1.4788	8.6992
0.0	6.0	3354	nan	1.5134	0.2001	1.4917	1.4788	8.6992
0.0	7.0	3913	nan	1.5134	0.2001	1.4917	1.4788	8.6992
0.0	8.0	4472	nan	1.5134	0.2001	1.4917	1.4788	8.6992
0.0	9.0	5031	nan	1.5134	0.2001	1.4917	1.4788	8.6992
0.0	10.0	5590	nan	1.5134	0.2001	1.4917	1.4788	8.6992
0.0	11.0	6149	nan	1.5134	0.2001	1.4917	1.4788	8.6992
0.0	12.0	6708	nan	1.5134	0.2001	1.4917	1.4788	8.6992
0.0	13.0	7267	nan	1.5134	0.2001	1.4917	1.4788	8.6992
0.0	14.0	7826	nan	1.5134	0.2001	1.4917	1.4788	8.6992
0.0	15.0	8385	nan	1.5134	0.2001	1.4917	1.4788	8.6992
0.0	16.0	8944	nan	1.5134	0.2001	1.4917	1.4788	8.6992
0.0	17.0	9503	nan	1.5134	0.2001	1.4917	1.4788	8.6992
0.0	18.0	10062	nan	1.5134	0.2001	1.4917	1.4788	8.6992
0.0	19.0	10621	nan	1.5134	0.2001	1.4917	1.4788	8.6992
0.0	20.0	11180	nan	1.5134	0.2001	1.4917	1.4788	8.6992
0.0	21.0	11739	nan	1.5134	0.2001	1.4917	1.4788	8.6992
0.0	22.0	12298	nan	1.5134	0.2001	1.4917	1.4788	8.6992
0.0	23.0	12857	nan	1.5134	0.2001	1.4917	1.4788	8.6992
0.0	24.0	13416	nan	1.5134	0.2001	1.4917	1.4788	8.6992
0.0	25.0	13975	nan	1.5134	0.2001	1.4917	1.4788	8.6992
0.0	26.0	14534	nan	1.5134	0.2001	1.4917	1.4788	8.6992
0.0	27.0	15093	nan	1.5134	0.2001	1.4917	1.4788	8.6992
0.0	28.0	15652	nan	1.5134	0.2001	1.4917	1.4788	8.6992
0.0	29.0	16211	nan	1.5134	0.2001	1.4917	1.4788	8.6992
0.0	30.0	16770	nan	1.5134	0.2001	1.4917	1.4788	8.6992
0.0	31.0	17329	nan	1.5134	0.2001	1.4917	1.4788	8.6992
0.0	32.0	17888	nan	1.5134	0.2001	1.4917	1.4788	8.6992
0.0	33.0	18447	nan	1.5134	0.2001	1.4917	1.4788	8.6992
0.0	34.0	19006	nan	1.5134	0.2001	1.4917	1.4788	8.6992
0.0	35.0	19565	nan	1.5134	0.2001	1.4917	1.4788	8.6992
0.0	36.0	20124	nan	1.5134	0.2001	1.4917	1.4788	8.6992
0.0	37.0	20683	nan	1.5134	0.2001	1.4917	1.4788	8.6992
0.0	38.0	21242	nan	1.5134	0.2001	1.4917	1.4788	8.6992
0.0	39.0	21801	nan	1.5134	0.2001	1.4917	1.4788	8.6992
0.0	40.0	22360	nan	1.5134	0.2001	1.4917	1.4788	8.6992
0.0	41.0	22919	nan	1.5134	0.2001	1.4917	1.4788	8.6992
0.0	42.0	23478	nan	1.5134	0.2001	1.4917	1.4788	8.6992
0.0	43.0	24037	nan	1.5134	0.2001	1.4917	1.4788	8.6992
0.0	44.0	24596	nan	1.5134	0.2001	1.4917	1.4788	8.6992
0.0	45.0	25155	nan	1.5134	0.2001	1.4917	1.4788	8.6992
0.0	46.0	25714	nan	1.5134	0.2001	1.4917	1.4788	8.6992
0.0	47.0	26273	nan	1.5134	0.2001	1.4917	1.4788	8.6992
0.0	48.0	26832	nan	1.5134	0.2001	1.4917	1.4788	8.6992
0.0	49.0	27391	nan	1.5134	0.2001	1.4917	1.4788	8.6992
0.0	50.0	27950	nan	1.5134	0.2001	1.4917	1.4788	8.6992

Framework versions

• Transformers 4.46.3
• Pytorch 2.5.1+cu118
• Datasets 3.1.0
• Tokenizers 0.20.3