bert-small2bert-small-finetuned-cnn_daily_mail-summarization-newsroom-filtered

This model is a fine-tuned version of mrm8488/bert-small2bert-small-finetuned-cnn_daily_mail-summarization on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 3.5413
• Rouge1: 32.3232
• Rouge2: 20.9203
• Rougel: 27.232
• Rougelsum: 29.345
• Gen Len: 72.2217

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: 4
• eval_batch_size: 4
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 500
• num_epochs: 5
• mixed_precision_training: Native AMP
• label_smoothing_factor: 0.1

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Gen Len
3.796	0.89	405	3.6945	29.7168	17.6705	24.4204	26.484	69.6847
3.6426	1.78	810	3.5532	32.3051	20.8789	27.1724	29.384	72.3695
3.2645	2.66	1215	3.5437	32.2016	20.758	27.083	29.0954	73.3892
3.1719	3.55	1620	3.5377	32.5493	21.083	27.0881	29.4691	71.5222
2.9763	4.44	2025	3.5413	32.3232	20.9203	27.232	29.345	72.2217

Framework versions

• Transformers 4.20.1
• Pytorch 1.11.0
• Datasets 2.1.0
• Tokenizers 0.12.1