---
language:
- en
- sql
license: mit
tags:
- text-to-sql
- agentic-data
- deepseek-r1
- llama-3
- data-engineering
datasets:
- custom-sql-corpus
metrics:
- execution-accuracy
pipeline_tag: text-generation
---

# Agentic Data 1

Agentic Data 1 is a specialized Large Language Model (LLM) fine-tuned for high-fidelity **Natural Language to SQL (NL2SQL)** generation. Built upon a proprietary high-performance foundation, it is designed for production data engineering workflows where output discipline and schema adherence are critical.

## 🚀 Model Capabilities

- **SQL Only Output**: Extremely high discipline in adhering to "SQL Only" instructions. No prose, no markdown, no reasoning leaks.
- **Complex Query Support**: Expert at handling multi-table joins, complex aggregations (Window functions, CTEs), and set operations.
- **Dialect Proficiency**: Optimized for PostgreSQL with broad compatibility for standard SQL dialects.
- **Enterprise Performance**: Delivers production-grade execution accuracy on complex enterprise-level SQL benchmarks.

## 🛠 Usage

### With vLLM (Recommended for Production)

```bash
python -m vllm.entrypoints.openai.api_server \
    --model DataManagement-AI/agentic-data-1-r3 \
    --served-model-name agentic-data-1-r3 \
    --max-model-len 8192
```

### With Ollama

```bash
ollama run DataManagement-AI/agentic-data-1-r3
```

| Benchmark | Metric | Status |
|---|---|---|
| Enterprise SQL Suite | Execution Accuracy | Verified |
| Output Discipline | Success Rate | 100% |

## ⚖️ License
This model is licensed under the MIT License.

## 🏢 About DataManagement.AI
DataManagement.AI is an organization dedicated to operationalizing AI for data engineering and governance.