metadata
tags:
- generated_from_trainer
datasets:
- samsum
metrics:
- rouge
model-index:
- name: pegasus-large-samsum
results:
- task:
name: Sequence-to-sequence Language Modeling
type: text2text-generation
dataset:
name: samsum
type: samsum
args: samsum
metrics:
- name: Rouge1
type: rouge
value: 48.0968
pegasus-large-samsum
This model is a fine-tuned version of google/pegasus-large on the samsum dataset. It achieves the following results on the evaluation set:
- Loss: 1.4109
- Rouge1: 48.0968
- Rouge2: 24.6663
- Rougel: 40.2569
- Rougelsum: 44.0137
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: 1
- eval_batch_size: 1
- seed: 42
- gradient_accumulation_steps: 64
- total_train_batch_size: 64
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 3
Training results
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum |
|---|---|---|---|---|---|---|---|
| No log | 1.0 | 230 | 1.4646 | 45.0631 | 22.5567 | 38.0518 | 41.2694 |
| No log | 2.0 | 460 | 1.4203 | 47.4122 | 24.158 | 39.7414 | 43.3485 |
| 1.699 | 3.0 | 690 | 1.4109 | 48.0968 | 24.6663 | 40.2569 | 44.0137 |
Framework versions
- Transformers 4.20.1
- Pytorch 1.12.0+cu113
- Datasets 2.3.2
- Tokenizers 0.12.1