| --- |
| license: apache-2.0 |
| library_name: peft |
| tags: |
| - trl |
| - sft |
| - generated_from_trainer |
| base_model: google/flan-t5-small |
| datasets: |
| - generator |
| model-index: |
| - name: output |
| results: [] |
| --- |
| |
| <!-- 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. --> |
|
|
| # output |
|
|
| This model is a fine-tuned version of [google/flan-t5-small](https://huggingface.co/google/flan-t5-small) on the generator dataset. |
|
|
| ## 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: 0.0002 |
| - train_batch_size: 4 |
| - eval_batch_size: 8 |
| - seed: 42 |
| - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 |
| - lr_scheduler_type: linear |
| - num_epochs: 1 |
|
|
| ### Framework versions |
|
|
| - PEFT 0.10.1.dev0 |
| - Transformers 4.39.1 |
| - Pytorch 2.2.1+cpu |
| - Datasets 2.18.0 |
| - Tokenizers 0.15.2 |
