metadata
library_name: transformers
license: apache-2.0
base_model: google/mt5-small
tags:
- generated_from_trainer
datasets:
- vitext2sql
metrics:
- bleu
model-index:
- name: TrainedModels
results:
- task:
name: Sequence-to-sequence Language Modeling
type: text2text-generation
dataset:
name: vitext2sql
type: vitext2sql
config: vitext2sql_source
split: test
args: vitext2sql_source
metrics:
- name: Bleu
type: bleu
value: 5.1741
TrainedModels
This model is a fine-tuned version of google/mt5-small on the vitext2sql dataset. It achieves the following results on the evaluation set:
- Loss: 0.7564
- Bleu: 5.1741
- Gen Len: 18.1829
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: 3
Training results
| Training Loss | Epoch | Step | Validation Loss | Bleu | Gen Len |
|---|---|---|---|---|---|
| 0.9517 | 1.0 | 6831 | 0.8409 | 3.1104 | 16.164 |
| 0.7029 | 2.0 | 13662 | 0.7696 | 4.9487 | 18.153 |
| 0.6078 | 3.0 | 20493 | 0.7564 | 5.1741 | 18.1829 |
Framework versions
- Transformers 4.44.2
- Pytorch 2.4.1+cu121
- Datasets 3.0.1
- Tokenizers 0.19.1