SummarEaseFocusV4

This model is a fine-tuned version of notBanana/SummarEaseFocusV3 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 1.3055
• Rouge1: 0.5919
• Rouge2: 0.4624
• Rougel: 0.5596
• Rougelsum: 0.5661
• Gen Len: 16.2667

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
• gradient_accumulation_steps: 16
• total_train_batch_size: 64
• 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	1	1.8613	0.3669	0.1974	0.3256	0.3281	20.0
No log	2.0	2	1.5717	0.4403	0.2774	0.404	0.406	19.6667
No log	3.0	3	1.3887	0.5442	0.3791	0.5119	0.5174	18.2667
No log	4.0	4	1.3055	0.5919	0.4624	0.5596	0.5661	16.2667

Framework versions

• Transformers 4.42.4
• Pytorch 2.3.1+cu121
• Datasets 2.20.0
• Tokenizers 0.19.1