metadata

license: mit
base_model: VietAI/vit5-large-vietnews-summarization
tags:
  - generated_from_trainer
metrics:
  - rouge
model-index:
  - name: mymodel_larger_vietnews_10k_2e5_3epoch_batchsize4
    results: []

mymodel_larger_vietnews_10k_2e5_3epoch_batchsize4

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

Loss: 1.9223
Rouge1: 0.5899
Rouge2: 0.281
Rougel: 0.3815
Rougelsum: 0.3817
Gen Len: 40.2035

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
num_epochs: 3
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Gen Len
1.8242	1.0	2000	1.7603	0.5792	0.2754	0.3786	0.3787	39.1905
1.3026	2.0	4000	1.8069	0.5884	0.2827	0.3827	0.3828	38.768
0.9538	3.0	6000	1.9223	0.5899	0.281	0.3815	0.3817	40.2035

Framework versions

Transformers 4.36.0.dev0
Pytorch 2.1.0+cu118
Datasets 2.15.0
Tokenizers 0.15.0