byt5-base-es_azz

This model is a fine-tuned version of google/byt5-base on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.6919
• Bleu: 5.6493
• Gen Len: 100.7258

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: 16
• eval_batch_size: 16
• seed: 65
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 10.0

Training results

Training Loss	Epoch	Step	Validation Loss	Bleu	Gen Len
No log	1.0	398	1.0151	0.1174	19.0
1.4581	2.0	796	0.8914	0.1875	19.0
1.0266	3.0	1194	0.8158	0.192	19.0
0.9168	4.0	1592	0.7706	0.2543	19.0
0.9168	5.0	1990	0.7434	0.3107	19.0
0.857	6.0	2388	0.7259	0.2904	19.0
0.8163	7.0	2786	0.7096	0.308	19.0
0.7885	8.0	3184	0.6969	0.2693	19.0
0.7718	9.0	3582	0.6932	0.3197	19.0
0.7718	10.0	3980	0.6919	0.2997	19.0

Framework versions

• Transformers 4.27.4
• Pytorch 2.0.0+cu118
• Datasets 2.11.0
• Tokenizers 0.13.3