Instructions to use cssupport/t5-small-awesome-text-to-sql with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use cssupport/t5-small-awesome-text-to-sql with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="cssupport/t5-small-awesome-text-to-sql")

# Load model directly
from transformers import AutoTokenizer, AutoModelForSeq2SeqLM

tokenizer = AutoTokenizer.from_pretrained("cssupport/t5-small-awesome-text-to-sql")
model = AutoModelForSeq2SeqLM.from_pretrained("cssupport/t5-small-awesome-text-to-sql")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use cssupport/t5-small-awesome-text-to-sql with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "cssupport/t5-small-awesome-text-to-sql"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "cssupport/t5-small-awesome-text-to-sql",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/cssupport/t5-small-awesome-text-to-sql

SGLang

How to use cssupport/t5-small-awesome-text-to-sql with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "cssupport/t5-small-awesome-text-to-sql" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "cssupport/t5-small-awesome-text-to-sql",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "cssupport/t5-small-awesome-text-to-sql" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "cssupport/t5-small-awesome-text-to-sql",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use cssupport/t5-small-awesome-text-to-sql with Docker Model Runner:
```
docker model run hf.co/cssupport/t5-small-awesome-text-to-sql
```

Training

by Raci - opened Apr 17, 2024

Discussion

Raci

Apr 17, 2024

Can you please share how to train this model with my own data set

Sayara

Apr 22, 2024

Hi,

I would like to retrain the model with my dataset too...Can you please provide the code for training?

roxitch

May 7, 2024

Hi,
Please can someone share the code to train this model with my database

igorChagas

Aug 30, 2024

Hello, may this article would be helpful "https://medium.com/@martinkeywood/fine-tuning-a-t5-small-model-to-generate-sql-from-natural-language-with-92-3-accuracy-fb29e062c638"

Kamal-Moha

Nov 9, 2024

Hi thanks, yes this article has been helpful to get things started.

I would like to know, how can I create a custom dataset to train a "Text to SQL" model. How can I create an acceptable custom dataset to proceed with training. Please help

mandartelitech

Jul 10, 2025

Use excel and pandas that will give you format to the same
import pandas as pd
import os
from google.colab import drive
drive.mount("/content/drive", force_remount=True)

file_path="/content/drive/My Drive/data1.csv"

print (os.path.isfile(file_path))

if os.path.exists(file_path):
print(f"File found at :{file_path}")
#!pip install -U datasets
from datasets import load_dataset , Dataset , DatasetDict
import pandas as pd

df=pd.read_csv(file_path)

#print(df)

#dataset = load_dataset("csv",data_files=file_path)
dataset= Dataset.from_pandas(df)

#print(dataset)

#for d in dataset:
#print(d)

dataset_dict= DatasetDict({"train":dataset})

#print(dataset_dict)

split_dataset=dataset_dict["train"].train_test_split(test_size=0.2)

#print(split_dataset)

#print(split_dataset["train"])
#print(split_dataset["test"])

train_dataset=split_dataset["train"]
test_dataset=split_dataset["test"]

my csv file looks like this may be you can create your own

role content
user What are the personal loans nos for last month
assistant get data from LOS request tabel loan_table
user What are new request for white card
assistant get data from credit_card database table credit_card_request
user What are the personal loans nos for last month
assistant get data from LOS request tabel loan_table
user What are new request for white card
assistant get data from credit_card database table credit_card_request
user What are the personal loans nos for last month
assistant get data from LOS request tabel loan_table
user What are new request for white card
assistant get data from credit_card database table credit_card_request
user What are the personal loans nos for last month
assistant get data from LOS request tabel loan_table
user What are new request for white card
assistant get data from credit_card database table credit_card_request
user What are the personal loans nos for last month
assistant get data from LOS request tabel loan_table
user What are new request for white card
assistant get data from credit_card database table credit_card_request
user What are the personal loans nos for last month
assistant get data from LOS request tabel loan_table

mandartelitech

Jul 10, 2025

Here do we need to provide prompt with create table everytime , I am looking for something where prompt will be natural language , trying build data extractor for enterprise

cssupport

Owner Jul 12, 2025

Yes you need to provide create table statement each time. You will need to train your model without create table statement.

Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.

Tap or paste here to upload images

· Sign up or log in to comment