YAML Metadata
Warning:
empty or missing yaml metadata in repo card
(https://huggingface.co/docs/hub/model-cards#model-card-metadata)
这个模型是根据这个一步一步完成的,如果想自己微调,请参考https://colab.research.google.com/github/huggingface/notebooks/blob/master/examples/summarization.ipynb This model is completed step by step according to this, if you want to fine-tune yourself, please refer to https://colab.research.google.com/github/huggingface/notebooks/blob/master/examples/summarization.ipynb
license: apache-2.0 tags: - generated_from_trainer datasets: - xsum metrics: - rouge model-index: - name: t5-small-finetuned-xsum results: - task: name: Sequence-to-sequence Language Modeling type: text2text-generation dataset: name: xsum type: xsum args: default metrics: - name: Rouge1 type: rouge value: 28.6901
t5-small-finetuned-xsum
This model is a fine-tuned version of t5-small on the xsum dataset. It achieves the following results on the evaluation set:
- Loss: 2.4500
- Rouge1: 28.6901
- Rouge2: 8.0102
- Rougel: 22.6087
- Rougelsum: 22.6105
- Gen Len: 18.824
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: 8
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 1
- mixed_precision_training: Native AMP
Training results
Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum | Gen Len |
---|---|---|---|---|---|---|---|---|
2.6799 | 1.0 | 25506 | 2.4500 | 28.6901 | 8.0102 | 22.6087 | 22.6105 | 18.824 |
Framework versions
- Transformers 4.12.3
- Pytorch 1.9.0+cu111
- Datasets 1.15.1
- Tokenizers 0.10.3
- Downloads last month
- 15
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social
visibility and check back later, or deploy to Inference Endpoints (dedicated)
instead.