model_v6

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: 0.5649
• Sacrebleu: 66.4771

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-06
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 16
• 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
0.5945	1.0	875	0.6579	63.2549
0.3681	2.0	1750	0.5965	65.0373
0.3426	3.0	2625	0.5690	65.4761
0.3215	4.0	3500	0.5700	65.8510
0.3084	5.0	4375	0.5637	66.2262
0.3051	6.0	5250	0.5657	66.4077
0.2908	7.0	6125	0.5682	66.4705
0.2868	8.0	7000	0.5661	66.4620
0.2795	9.0	7875	0.5761	66.5109
0.2832	10.0	8750	0.5649	66.4771

Framework versions

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