bart-cnn-science-v3-e4

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

• Loss: 0.8265
• Rouge1: 53.0296
• Rouge2: 33.4957
• Rougel: 35.8876
• Rougelsum: 50.0786
• Gen Len: 141.5926

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: 2
• eval_batch_size: 2
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 4
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Gen Len
No log	1.0	398	0.9965	52.4108	32.1506	35.0281	50.0368	142.0
1.176	2.0	796	0.8646	52.7182	32.9681	35.1454	49.9527	141.8333
0.7201	3.0	1194	0.8354	52.5417	32.6428	35.8703	49.8037	142.0
0.5244	4.0	1592	0.8265	53.0296	33.4957	35.8876	50.0786	141.5926

Framework versions

• Transformers 4.19.2
• Pytorch 1.11.0+cu113
• Datasets 2.2.2
• Tokenizers 0.12.1