emnlp_test_9clusters_msmarco_bart_large_200k

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: 0.6447
• Bleu: 77.0949
• Gen Len: 10.5172

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: 128
• eval_batch_size: 128
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 6
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Bleu	Gen Len
1.8802	1.0	1250	1.1652	53.9226	10.7576
0.8808	2.0	2500	0.8230	68.163	10.5673
0.6896	3.0	3750	0.7116	73.3521	10.6162
0.5654	4.0	5000	0.7829	75.4941	10.449
0.5079	5.0	6250	0.6514	76.6795	10.5106
0.4738	6.0	7500	0.6447	77.0949	10.5172

Framework versions

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