|
--- |
|
license: apache-2.0 |
|
base_model: google/flan-t5-small |
|
tags: |
|
- generated_from_trainer |
|
model-index: |
|
- name: text-correction-en-small |
|
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. --> |
|
|
|
# text-correction-en-small |
|
|
|
This model is a fine-tuned version of [google/flan-t5-small](https://huggingface.co/google/flan-t5-small) on the small portion of |
|
this [dataset](https://huggingface.co/datasets/gudleifrr/text-correction-en). |
|
|
|
## Model description |
|
|
|
This model is fine-tuned on the task of text correction: it fixes English text with the wrong grammar, punctuation and order of words. |
|
Significantly larger model that supports Russian language as well can be found [here](https://huggingface.co/gudleifrr/mt5-text-correction-enru). |
|
|
|
|
|
## Training and evaluation data |
|
|
|
Around 10% of the data from this [dataset](https://huggingface.co/datasets/gudleifrr/text-correction-en) was used. |
|
|
|
|
|
### Training hyperparameters |
|
|
|
The following hyperparameters were used during training: |
|
- learning_rate: 5e-05 |
|
- train_batch_size: 4 |
|
- eval_batch_size: 4 |
|
- seed: 42 |
|
- gradient_accumulation_steps: 2 |
|
- total_train_batch_size: 8 |
|
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 |
|
- lr_scheduler_type: linear |
|
- lr_scheduler_warmup_steps: 100 |
|
- num_epochs: 1 |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.41.0 |
|
- Pytorch 2.2.2 |
|
- Datasets 2.19.1 |
|
- Tokenizers 0.19.1 |
|
|
