metadata

license: apache-2.0
tags:
  - generated_from_trainer
datasets:
  - farleyknight/big_patent_5_percent
metrics:
  - rouge
model-index:
  - name: patent-summarization-allen-led-large-2022-09-20
    results:
      - task:
          name: Summarization
          type: summarization
        dataset:
          name: farleyknight/big_patent_5_percent
          type: farleyknight/big_patent_5_percent
          config: all
          split: train
          args: all
        metrics:
          - name: Rouge1
            type: rouge
            value: 0

patent-summarization-allen-led-large-2022-09-20

This model is a fine-tuned version of allenai/led-large-16384-arxiv on the farleyknight/big_patent_5_percent dataset. It achieves the following results on the evaluation set:

Loss: 2.8233
Rouge1: 0.0
Rouge2: 0.0
Rougel: 0.0
Rougelsum: 0.0
Gen Len: 128.0

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: 1
eval_batch_size: 1
seed: 42
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
num_epochs: 1.0

Training results

Training Loss	Epoch	Step	Validation Loss	Gen Len
3.4766	0.08	5000	3.4240	512.0
3.2549	0.17	10000	3.2908	512.0
3.2295	0.25	15000	3.1862	512.0
3.1455	0.33	20000	3.1291	512.0
3.0526	0.41	25000	3.0684	512.0
3.0024	0.5	30000	3.0134	512.0
2.9671	0.58	35000	2.9696	512.0
2.9862	0.66	40000	2.9431	512.0
2.9168	0.75	45000	2.8989	512.0
2.9063	0.83	50000	2.8559	512.0
2.8417	0.91	55000	2.8398	512.0
2.7853	0.99	60000	2.8240	512.0

Framework versions

Transformers 4.23.0.dev0
Pytorch 1.12.0
Datasets 2.4.0
Tokenizers 0.12.1