metadata
library_name: transformers
license: apache-2.0
base_model: google/mt5-large
tags:
- generated_from_trainer
metrics:
- rouge
model-index:
- name: mt5-large-ie-budquo-5k
results: []
mt5-large-ie-budquo-5k
This model is a fine-tuned version of google/mt5-large on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.0504
- Rouge1: 0.1024
- Rouge2: 0.0848
- Rougel: 0.1021
- Rougelsum: 0.1023
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: 0.0003
- train_batch_size: 1
- eval_batch_size: 1
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 4
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 5
Training results
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum |
|---|---|---|---|---|---|---|---|
| 0.1081 | 0.9997 | 1921 | 0.0537 | 0.0995 | 0.0819 | 0.0994 | 0.0993 |
| 0.0291 | 2.0 | 3843 | 0.0177 | 0.1030 | 0.0866 | 0.1030 | 0.1029 |
| 0.0157 | 2.9997 | 5764 | 0.0101 | 0.1041 | 0.0878 | 0.1040 | 0.1037 |
| 0.0092 | 4.0 | 7686 | 0.0071 | 0.1039 | 0.0881 | 0.1040 | 0.1038 |
| 0.0061 | 4.9987 | 9605 | 0.0067 | 0.1045 | 0.0884 | 0.1043 | 0.1041 |
Framework versions
- Transformers 4.44.2
- Pytorch 2.4.1+cu121
- Datasets 3.0.1
- Tokenizers 0.19.1