eslamxm's picture
update model card README.md
e4146c3
|
raw
history blame
1.42 kB
metadata
license: apache-2.0
tags:
  - summarization
  - mt5
  - es
  - spanish
  - abstractive summarization
  - generated_from_trainer
datasets:
  - wiki_lingua
model-index:
  - name: mt5-base-finetuned-Spanish
    results: []

mt5-base-finetuned-Spanish

This model is a fine-tuned version of google/mt5-base on the wiki_lingua dataset. It achieves the following results on the evaluation set:

  • Loss: 3.1727
  • Rouge-1: 28.11
  • Rouge-2: 12.09
  • Rouge-l: 24.62
  • Gen Len: 18.73
  • Bertscore: 72.25

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: 0.0005
  • train_batch_size: 4
  • eval_batch_size: 4
  • seed: 42
  • gradient_accumulation_steps: 8
  • total_train_batch_size: 32
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • lr_scheduler_warmup_steps: 250
  • num_epochs: 5
  • label_smoothing_factor: 0.1

Training results

Framework versions

  • Transformers 4.19.4
  • Pytorch 1.11.0+cu113
  • Datasets 2.3.0
  • Tokenizers 0.12.1