Instructions to use JohnnyZeppelin/sql-ocr with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use JohnnyZeppelin/sql-ocr with Transformers:
# Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("JohnnyZeppelin/sql-ocr", dtype="auto") - Notebooks
- Google Colab
- Kaggle
This folder contains the lightweight table-to-SQL adapter trained locally in this task.
Files:
model.safetensors.base64: base64-encoded weights (decode withbase64 -d model.safetensors.base64 > model.safetensors).- Tokenizer/config files copied from the training output to load the model after decoding.
Command to reconstruct the weights:
base64 -d model.safetensors.base64 > model.safetensors