mt5_base_EN_TH_sch_wiki

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
• Rouge2 Precision: 0.0098
• Rouge2 Recall: 0.0051
• Rouge2 Fmeasure: 0.0065

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: 45
• eval_batch_size: 16
• seed: 42
• optimizer: Use 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: 15
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge2 Precision	Rouge2 Recall	Rouge2 Fmeasure
0.0	1.0	2879	nan	0.0098	0.0051	0.0065
0.0	2.0	5758	nan	0.0098	0.0051	0.0065
0.0	3.0	8637	nan	0.0098	0.0051	0.0065
0.0	4.0	11516	nan	0.0098	0.0051	0.0065
0.0	5.0	14395	nan	0.0098	0.0051	0.0065
0.0	6.0	17274	nan	0.0098	0.0051	0.0065
0.0	7.0	20153	nan	0.0098	0.0051	0.0065
0.0	8.0	23032	nan	0.0098	0.0051	0.0065
0.0	9.0	25911	nan	0.0098	0.0051	0.0065
0.0	10.0	28790	nan	0.0098	0.0051	0.0065
0.0	11.0	31669	nan	0.0098	0.0051	0.0065
0.0	12.0	34548	nan	0.0098	0.0051	0.0065
0.0	13.0	37427	nan	0.0098	0.0051	0.0065
0.0	14.0	40306	nan	0.0098	0.0051	0.0065
0.0	15.0	43185	nan	0.0098	0.0051	0.0065

Framework versions

• Transformers 4.46.2
• Pytorch 2.2.2
• Datasets 2.16.1
• Tokenizers 0.20.3