Spaces:
Sleeping
Sleeping
Commit History
-
Delete sensitive auth_db.json
DvwnUnverified signaturea652cd8
-
Update UI Frontend version 1.2.0 + Backend Integration
a5b3518
-
Update UI Frontend Version 1.1.0
a099ae2
-
Streamlit UI Frontend version 1.0.0
723d980 -
Frontend Structure and Framework Setup
b33b249
-
Delete App.vue
1e2664e -
Update README.md
3161562 -
Update main.py and README.md version 1.1.0
- Update menu and selected model - Create a README.md733a707
-
Create Q&A Report Evaluation
a5862c5 -
Update evaluation_mode.py & interactive_mode.py version 1.1.0
- Update scripts to automate the testing and evaluation. - Write into file for comparison.c96208b
-
hf_engine.py & sql_agent.py version 1.1.0
Updated the function for model selectione06da36
-
Debugging
- Remove duplicate & unnecessary filesa4607e0 -
Merge branch 'main' into NL2SQL-V1.0
f3f9320 -
Folder and Files Structure Refactor
- Adding backend and frontend files31c1d8c -
Update evaluation mode
- To evaluate which categories causes the low ex and esm. - Adding new method for model registry on hf_engine.py file - Adding models for testingdfa643b
-
Update interactive mode testing module
1. Show results/data return by agent in a tabular format and summarized sentences of the answer 2. Using tabulate lib packagef160d1e
-
Delete unused files
91c6f8c -
Update Agent Response using NL
Extract and Return data in NLedef444 -
Merge branch 'main' of https://github.com/Dvwn00/NL2SQL
7ce8a1d -
Update Agent response in NL
1. Update main.py to separate the interactive logic keeping it modular. 2. Create new file (interactive_mode.py) to handle conversational mode. 3. Update sql_agent.py to include NL response template so system return NL response to user with data needed instead of only generated SQL. 4. Update evaluate_hf.py -> evaluation_mode.py to make naming convention consistent.9f8f0ff
-
Delete freeze
DvwnUnverified signature9ace1d8 -
Update qwen_engine.py comment
d9b3c34 -
Update Engine Model and HF Token
1. Currently, change SQLCoder to Qwen2.5-Coder. 2. Add token loader on main entry application, main.py.8c0edc5 -
Update SQL Agent Auto-Correct Mechanism
79168b4 -
Update main entry point for nl2sql application > app/main.py
1. Run from the main entry point, main.py 2. Remove local test block for db_manager, hf_engine, and sql_agenta218c12 -
Update evaluation script and test cases json
1. Update evaluation scripts to utilise agent sql instead of directly use engine LLM 2. Tweak questions and sql in test cases json71c851f
-
Update SQL agent > src/nl2sql/sql_agent.py
1. Act as a bridge connect the db_manager and hf_engine. 2. Utilize PromptTemplate to instruct LLM on its persona.dc6fe88 -
Update HF Model Engine > hf_engine.py
1. Implement custom wrapper for LangChain compatibility for later agent implementation. 2. Use InferenceClient to initialize the model with Inference API Provider 3. Tested with local code block - Successfully loaded5ac1ec2 -
Update Schema Retrieval using RAG and LangChain
1. Packages Requirements: pip install > langchain-community chromadb langchain-chroma langchain-huggingface sentence-transformer 2. File Changes: - db_manager.py - .env - .gitignore - requirements.txtccb82de
-
Try and Error to Improve Accuracy
Trying Codex to improve model accuracy0731ede
-
NL2SQL Qwen Model V1.0.0
97dafec
-
Initial commit
DvwnUnverified signature9ef286c