Instructions to use juierror/flan-t5-text2sql-with-schema-v2 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use juierror/flan-t5-text2sql-with-schema-v2 with Transformers:

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

pipe = pipeline("text-generation", model="juierror/flan-t5-text2sql-with-schema-v2")

# Load model directly
from transformers import AutoTokenizer, AutoModelForSeq2SeqLM

tokenizer = AutoTokenizer.from_pretrained("juierror/flan-t5-text2sql-with-schema-v2")
model = AutoModelForSeq2SeqLM.from_pretrained("juierror/flan-t5-text2sql-with-schema-v2")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use juierror/flan-t5-text2sql-with-schema-v2 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "juierror/flan-t5-text2sql-with-schema-v2"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "juierror/flan-t5-text2sql-with-schema-v2",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/juierror/flan-t5-text2sql-with-schema-v2

SGLang

How to use juierror/flan-t5-text2sql-with-schema-v2 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 "juierror/flan-t5-text2sql-with-schema-v2" \
    --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": "juierror/flan-t5-text2sql-with-schema-v2",
		"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 "juierror/flan-t5-text2sql-with-schema-v2" \
        --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": "juierror/flan-t5-text2sql-with-schema-v2",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use juierror/flan-t5-text2sql-with-schema-v2 with Docker Model Runner:
```
docker model run hf.co/juierror/flan-t5-text2sql-with-schema-v2
```

Calculating Accuracy of the model

by ahmedrizwan239 - opened Feb 29, 2024

Discussion

ahmedrizwan239

Feb 29, 2024

Since I'm new into NLP, idk how to calculate the accuracy of the model you've developed.

I got your code from github, trained it on the same dataset and I think the validation loss is greater.

Please guide me if you can, TIA!

ahmedrizwan239

Feb 29, 2024

I generated this code below:

from sklearn.metrics import accuracy_score

predictions = trainer.predict(val_dataset).predictions
references = val_dataset ["query"]
accuracy = accuracy_ score(references, predictions)
print(f"Accuracy: {accuracy: .4f}")

Do you think it should work fine? Please let me know if there are any suggestions or modifications required. Thanks!

juierror

Owner Feb 29, 2024

@ahmedrizwan239 Hi

since this is a text generative task, I think bleu score is more appropriate than accuracy

you can check this for more information
https://huggingface.co/spaces/evaluate-metric/bleu

ahmedrizwan239

Feb 29, 2024

Thanks a lot. You are right, bleu score is more suitable for generative tasks.

I calculated the bleu score and rouge score, here are the results:
{'bleu': 69.11695994149133, 'rouge1': 93.32338243439364, 'rouge2': 87.46108869509003, 'rougeL': 91.20303118257938}

I think these are pretty good.

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