https://github.com/hetpandya/paraphrase-datasets-pretrained-models
A collection of preprocessed datasets and pretrained models for generating paraphrases.
https://github.com/hetpandya/paraphrase-datasets-pretrained-models
Last synced: 3 months ago
JSON representation
A collection of preprocessed datasets and pretrained models for generating paraphrases.
- Host: GitHub
- URL: https://github.com/hetpandya/paraphrase-datasets-pretrained-models
- Owner: hetpandya
- License: apache-2.0
- Created: 2021-06-28T08:31:49.000Z (almost 4 years ago)
- Default Branch: main
- Last Pushed: 2021-07-13T12:44:52.000Z (almost 4 years ago)
- Last Synced: 2025-01-15T21:30:36.875Z (4 months ago)
- Homepage:
- Size: 5.21 MB
- Stars: 29
- Watchers: 1
- Forks: 4
- Open Issues: 1
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# Paraphrase datasets and pretrained models
This repository consists of a collection of preprocessed datasets for training models and pretrained models to generate paraphrases.
## Datasets
Each dataset has a `README` that describes the dataset, its source and preprocessed format. The datasets are stored in the [datasets](https://github.com/hetpandya/paraphrase-datasets-pretrained-models/tree/main/datasets) directory.
Dataset Type| File Name
--- | ---
TaPaCo Original| [tapaco_huggingface.csv](https://github.com/hetpandya/paraphrase-datasets-pretrained-models/blob/main/datasets/tapaco/tapaco_huggingface.csv)
TaPaCo Preprocessed| [tapaco_paraphrases_dataset.csv](https://github.com/hetpandya/paraphrase-datasets-pretrained-models/blob/main/datasets/tapaco/tapaco_paraphrases_dataset.csv)
## Pretrained models
List of models trained on various datasets for paraphrase generation.
Model| Dataset | Location
--- | --- | --- |
t5-small| [tapaco](https://github.com/hetpandya/paraphrase-datasets-pretrained-models/tree/main/datasets/tapaco) | [huggingface](https://huggingface.co/hetpandya/t5-small-tapaco)
t5-small| [Quora Question Pairs](https://quoradata.quora.com/First-Quora-Dataset-Release-Question-Pairs) | [huggingface](https://huggingface.co/hetpandya/t5-small-quora)
t5-base| [tapaco](https://github.com/hetpandya/paraphrase-datasets-pretrained-models/tree/main/datasets/tapaco) | [huggingface](https://huggingface.co/hetpandya/t5-base-tapaco)
### Model Training
Examples for training models on the datasets can be found in the [examples](https://github.com/hetpandya/paraphrase-datasets-pretrained-models/tree/main/examples) directory.
### T5 model usage example
**Install dependencies using:**
```
pip install transformers sentencepiece
```
**Usage**
```python
from transformers import T5ForConditionalGeneration, T5Tokenizer
tokenizer = T5Tokenizer.from_pretrained("hetpandya/t5-small-tapaco")
model = T5ForConditionalGeneration.from_pretrained("hetpandya/t5-small-tapaco")
def get_paraphrases(sentence, prefix="paraphrase: ", n_predictions=5, top_k=120, max_length=256,device="cpu"):
text = prefix + sentence + " "
encoding = tokenizer.encode_plus(
text, pad_to_max_length=True, return_tensors="pt"
)
input_ids, attention_masks = encoding["input_ids"].to(device), encoding[
"attention_mask"
].to(device)
model_output = model.generate(
input_ids=input_ids,
attention_mask=attention_masks,
do_sample=True,
max_length=max_length,
top_k=top_k,
top_p=0.98,
early_stopping=True,
num_return_sequences=n_predictions,
)
outputs = []
for output in model_output:
generated_sent = tokenizer.decode(
output, skip_special_tokens=True, clean_up_tokenization_spaces=True
)
if (
generated_sent.lower() != sentence.lower()
and generated_sent not in outputs
):
outputs.append(generated_sent)
return outputs
paraphrases = get_paraphrases("The house will be cleaned by me every Saturday.")
for sent in paraphrases:
print(sent)
```
**Output**
```
The house is cleaned every Saturday by me.
The house will be cleaned on Saturday.
I will clean the house every Saturday.
I get the house cleaned every Saturday.
I will clean this house every Saturday.
```