pegasus-cnn_dailymail-v4-e1-e4-feedback

This model is a fine-tuned version of theojolliffe/pegasus-cnn_dailymail-v4-e1 on the None dataset. It achieves the following results on the evaluation set:

• Loss: 5.1815
• Rouge1: 59.7329
• Rouge2: 42.429
• Rougel: 53.5945
• Rougelsum: 52.9219
• Gen Len: 37.8636

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

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Gen Len
No log	1.0	90	5.5674	60.4652	43.8057	54.5759	52.5243	39.5
No log	2.0	180	5.3261	60.248	43.9953	54.9492	53.6372	38.6364
No log	3.0	270	5.2167	60.4887	43.6915	54.3825	53.5206	37.1364
No log	4.0	360	5.1815	59.7329	42.429	53.5945	52.9219	37.8636

Framework versions

• Transformers 4.12.3
• Pytorch 1.9.0
• Datasets 1.18.0
• Tokenizers 0.10.3