model

This model is a fine-tuned version of facebook/bart-large on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.6118
• Sacrebleu: 67.0674

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: 1e-05
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 64
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 10
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Sacrebleu
No log	1.0	218	0.5748	65.7729
No log	2.0	437	0.5609	66.4458
No log	3.0	656	0.5786	66.8800
No log	4.0	875	0.5485	66.8774
No log	5.0	1093	0.5506	66.8932
No log	6.0	1312	0.5890	66.9609
No log	7.0	1531	0.5674	66.8977
No log	8.0	1750	0.5903	66.9902
No log	9.0	1968	0.6196	67.0418
No log	9.97	2180	0.6118	67.0674

Framework versions

• Transformers 4.39.3
• Pytorch 2.1.2
• Datasets 2.18.0
• Tokenizers 0.15.2