|
|
--- |
|
|
language: en |
|
|
tags: |
|
|
- text2sql |
|
|
- causal-lm |
|
|
- transformers |
|
|
- safetensors |
|
|
license: mit |
|
|
pipeline_tag: text-generation |
|
|
--- |
|
|
|
|
|
|
|
|
|
|
|
# Phi-3 Text-to-SQL Model |
|
|
|
|
|
This is a fine-tuned **Microsoft Phi-3** model specialized for **Text-to-SQL** generation. |
|
|
|
|
|
## Example |
|
|
|
|
|
```python |
|
|
from transformers import AutoModelForCausalLM, AutoTokenizer |
|
|
|
|
|
repo_id = "bhavika24/text2sql" |
|
|
tokenizer = AutoTokenizer.from_pretrained(repo_id) |
|
|
model = AutoModelForCausalLM.from_pretrained(repo_id, torch_dtype="auto", device_map="auto") |
|
|
|
|
|
question = "List all customers who ordered products over $500 last month." |
|
|
inputs = tokenizer(question, return_tensors="pt").to(model.device) |
|
|
outputs = model.generate(**inputs, max_new_tokens=128) |
|
|
|
|
|
sql_query = tokenizer.decode(outputs[0], skip_special_tokens=True) |
|
|
print(sql_query) |
|
|
|
