<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->
flan-t5-large-extraction-cnndm_8000-all-new
This model is a fine-tuned version of google/flan-t5-large on the None dataset. It achieves the following results on the evaluation set:
- Loss: 1.6853
- Rouge1: 34.5211
- Rouge2: 15.298
- Rougel: 29.4229
- Rougelsum: 29.5024
- Gen Len: 19.0
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: 2e-05
- train_batch_size: 16
- eval_batch_size: 48
- seed: 1799
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 20
Training results
Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum | Gen Len |
---|---|---|---|---|---|---|---|---|
2.1308 | 0.4 | 200 | 1.7871 | 34.3185 | 14.8966 | 29.6706 | 29.7468 | 19.0 |
1.9291 | 0.8 | 400 | 1.7315 | 34.4255 | 15.1159 | 29.5114 | 29.609 | 18.99 |
1.8624 | 1.2 | 600 | 1.7356 | 35.4244 | 15.3865 | 30.0524 | 30.0961 | 18.99 |
1.8106 | 1.6 | 800 | 1.7151 | 34.8718 | 15.5046 | 29.9394 | 30.0119 | 19.0 |
1.7948 | 2.0 | 1000 | 1.7103 | 33.8176 | 14.974 | 29.3676 | 29.4451 | 18.99 |
1.7303 | 2.4 | 1200 | 1.6940 | 34.8851 | 15.4393 | 30.1554 | 30.2106 | 19.0 |
1.7276 | 2.8 | 1400 | 1.6853 | 34.5211 | 15.298 | 29.4229 | 29.5024 | 19.0 |
1.6804 | 3.2 | 1600 | 1.7031 | 35.0279 | 15.7442 | 29.9333 | 30.0312 | 19.0 |
1.6597 | 3.6 | 1800 | 1.6939 | 35.4139 | 15.9005 | 30.1214 | 30.1721 | 18.998 |
1.6432 | 4.0 | 2000 | 1.6874 | 34.8249 | 15.4814 | 29.794 | 29.8873 | 18.998 |
1.5921 | 4.4 | 2200 | 1.6940 | 34.7792 | 15.5349 | 29.6781 | 29.7601 | 19.0 |
1.6044 | 4.8 | 2400 | 1.6956 | 35.7927 | 16.1867 | 30.6689 | 30.7588 | 19.0 |
Framework versions
- Transformers 4.18.0
- Pytorch 1.10.0+cu111
- Datasets 2.5.1
- Tokenizers 0.12.1