results

This model is a fine-tuned version of google/roberta2roberta_L-24_cnn_daily_mail on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 6.7395
• Rouge1: 33.11
• Rouge2: 20.39
• Rougel: 27.32
• Rougelsum: 27.42
• Bertscore P: 87.57
• Bertscore R: 83.35
• Bertscore F1: 85.27

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: 3e-05
• train_batch_size: 2
• eval_batch_size: 2
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 8
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: cosine
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 3

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Bertscore P	Bertscore R	Bertscore F1
36.5558	0.8	20	7.8969	32.35	20.37	27.74	27.82	87.74	83.51	85.43
32.2661	1.6	40	7.0747	34.94	21.86	29.83	30.04	87.7	83.93	85.62
30.3129	2.4	60	6.7395	33.11	20.39	27.32	27.42	87.57	83.35	85.27

Framework versions

• PEFT 0.14.0
• Transformers 4.48.3
• Pytorch 2.5.1+cu124
• Datasets 3.3.0
• Tokenizers 0.21.0