bart-large-translation-spa-guc

This model is a fine-tuned version of facebook/bart-large on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.8484
• Bleu: 3.2289
• Gen Len: 18.2771

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: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 10
• num_epochs: 15
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Bleu	Gen Len
1.5556	1.0	7668	1.2930	1.2368	18.9444
1.4604	2.0	15336	1.0780	1.9761	18.4493
0.8076	3.0	23004	0.9928	2.2387	18.297
1.1913	4.0	30672	0.9398	2.6084	18.2087
0.8532	5.0	38340	0.8947	2.5809	18.2469
0.6234	6.0	46008	0.8649	2.7376	18.2842
0.7989	7.0	53676	0.8535	2.8415	18.2283
0.6287	8.0	61344	0.8512	2.9061	18.1744
0.7429	9.0	69012	0.8471	2.9767	18.4115
0.3585	10.0	76680	0.8442	3.1551	18.2665
1.0195	11.0	84348	0.8484	3.2289	18.2771

Framework versions

• Transformers 4.35.2
• Pytorch 1.13.1+cu117
• Datasets 2.15.0
• Tokenizers 0.15.0