vit5-base-vietnews-summarization-finetuned-VN

This model is a fine-tuned version of VietAI/vit5-base-vietnews-summarization on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 2.2948
• Rouge1: 44.6941
• Rouge2: 20.3126
• Rougel: 31.4356
• Rougelsum: 32.0879
• Gen Len: 18.9868

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: 4
• total_train_batch_size: 16
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 5

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Gen Len
No log	1.0	266	2.3453	44.2095	19.8671	31.0644	31.688	18.9519
2.5909	2.0	533	2.2875	44.4455	20.2345	31.3714	31.9954	18.97
2.5909	3.0	799	2.2792	44.4222	20.1061	31.2145	31.8668	18.9791
1.9843	4.0	1066	2.2806	44.6434	20.2502	31.3347	31.9962	18.9833
1.9843	4.99	1330	2.2948	44.6941	20.3126	31.4356	32.0879	18.9868

Framework versions

• Transformers 4.30.0
• Pytorch 2.3.0+cu121
• Datasets 2.20.0
• Tokenizers 0.13.3