umt5-base-quechua-espanol-finetuned-model-v3

This model is a fine-tuned version of google/umt5-base on the None dataset. It achieves the following results on the evaluation set:

• Loss: 1.4864

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: 0.0002
• train_batch_size: 22
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: constant
• num_epochs: 2

Training results

Training Loss	Epoch	Step	Validation Loss
3.0774	0.2251	1000	2.5173
2.5309	0.4502	2000	2.1156
2.2481	0.6754	3000	1.8950
2.065	0.9005	4000	1.7553
1.8966	1.1256	5000	1.6603
1.8182	1.3507	6000	1.5891
1.762	1.5759	7000	1.5303
1.7083	1.8010	8000	1.4864

Framework versions

• Transformers 4.41.2
• Pytorch 2.3.0+cu121
• Datasets 2.20.0
• Tokenizers 0.19.1