md_mt5_0109

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

• Loss: 0.4790
• Bleu: 0.457
• Gen Len: 18.9295

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: 2e-05
• train_batch_size: 4
• eval_batch_size: 4
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 15

Training results

Training Loss	Epoch	Step	Validation Loss	Bleu	Gen Len
13.8417	1.0	975	2.6438	0.563	15.6487
2.8117	2.0	1950	1.4148	0.891	17.2223
1.8883	3.0	2925	1.0693	0.401	18.7582
1.5248	4.0	3900	0.8703	0.4583	18.8508
1.3116	5.0	4875	0.7483	0.4651	18.8856
1.1617	6.0	5850	0.6783	0.4542	18.9005
1.0636	7.0	6825	0.6243	0.459	18.9054
0.9928	8.0	7800	0.5869	0.4707	18.9038
0.9272	9.0	8775	0.5536	0.4563	18.9031
0.8926	10.0	9750	0.5282	0.4606	18.9177
0.8568	11.0	10725	0.5091	0.4577	18.9226
0.8341	12.0	11700	0.4964	0.4482	18.9259
0.8176	13.0	12675	0.4867	0.4539	18.9262
0.806	14.0	13650	0.4812	0.4576	18.9264
0.7945	15.0	14625	0.4790	0.457	18.9295

Framework versions

• Transformers 4.35.2
• Pytorch 2.1.0+cu121
• Datasets 2.16.1
• Tokenizers 0.15.0