metadata

license: apache-2.0
language:
  - en
tags:
  - generated_from_trainer
model-index:
  - name: t5-small-finetuned-turk-text-simplification
    results: []
widget:
  - text: >-
      simplify: the incident has been the subject of numerous reports as to
      ethics in scholarship .
  - text: >-
      simplify: the historical method comprises the techniques and guidelines by
      which historians use primary sources and other evidence to research and
      then to write history .
  - text: >-
      simplify: none of the authors , contributors , sponsors , administrators ,
      vandals , or anyone else connected with wikipedia , in any way whatsoever
      , can be responsible for your use of the information contained in or
      linked from these web pages .
  - text: 'simplify: oregano is an indispensable ingredient in greek cuisine .'
inference:
  parameters:
    temperature: 1.5
    max_length: 256
    do_sample: true
    num_beams: 3

T5 (small) finetuned-turk-text-simplification

This model is a fine-tuned version of t5-small on the None dataset. It achieves the following results on the evaluation set:

Loss: 0.1001
Rouge2 Precision: 0.6825
Rouge2 Recall: 0.4542
Rouge2 Fmeasure: 0.5221

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: 5e-05
train_batch_size: 32
eval_batch_size: 32
seed: 42
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
num_epochs: 5
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge2 Precision	Rouge2 Recall	Rouge2 Fmeasure
0.4318	1.0	500	0.1053	0.682	0.4533	0.5214
0.0977	2.0	1000	0.1019	0.683	0.4545	0.5225
0.0938	3.0	1500	0.1010	0.6828	0.4547	0.5226
0.0916	4.0	2000	0.1003	0.6829	0.4545	0.5225
0.0906	5.0	2500	0.1001	0.6825	0.4542	0.5221

Framework versions

Transformers 4.21.3
Pytorch 1.12.1+cu113
Datasets 2.4.0
Tokenizers 0.12.1