# Grape Models — NLP to Terminal

> A collection of T5 expert models hosted on Hugging Face focused on translating natural language into actionable terminal commands, routed by an intelligent semantic categorizer.

**Live:** [grape.jrodriiguezg.link](https://grape.jrodriiguezg.link)  
**Collection:** [huggingface.co/collections/jrodriiguezg/grape-models](https://huggingface.co/collections/jrodriiguezg/grape-models)

---

## Overview

Grape Models is a **Mixture of Experts (MoE)** NLP system that translates natural language instructions into terminal commands. It consists of four specialized expert models — each fine-tuned for a specific domain of terminal operations — and a semantic router that directs each query to the correct expert.

## Architecture

```
User Input (Natural Language)
         │
         ▼
  ┌─────────────────┐
  │  Semantic Router │  ← minilm-l12-grape-route
  └─────────────────┘
         │
    ┌────┴────────────────────┐
    ▼        ▼        ▼       ▼
  BASH    SEARCH   DOCKER  NETWORK
  Expert  Expert   Expert  Expert
    │        │        │       │
    └────────┴────────┴───────┘
                 │
          Terminal Command
```

### Base Foundation

All 4 expert models share the same base: **`Salesforce/codet5-small`**, enabling lightweight execution without sacrificing quality.

### How it works

```python
# NLP Pipeline Concept
def process_input(text):
    # 1. Router categorizes intent
    category = router_model.predict(text)
    
    # 2. Select specific Grape Expert
    expert = load_expert(category)
    
    # 3. Generate terminal command
    command = expert.generate(text)
    return command
```

## Expert Models

| Model | Domain | Description | Link |
|---|---|---|---|
| **Chardonnay** | `BASH` | Generalist — file management, networking, system operations | [Model Card →](https://huggingface.co/jrodriiguezg/grape-chardonnay) |
| **Pinot** | `SEARCH` | Search & retrieval with Zero-Modification Policy (`find`, `grep`, `stat`) | [Model Card →](https://huggingface.co/jrodriiguezg/grape-pinot) |
| **Malbec** | `DOCKER` | Docker & docker-compose workflows | [Model Card →](https://huggingface.co/jrodriiguezg/grape-malbec) |
| **Syrah** | `NETWORK` | Network tasks — port scanning, DNS, connectivity, firewall | [Model Card →](https://huggingface.co/jrodriiguezg/grape-syrah) |

### Semantic Router

| Model | Purpose | Link |
|---|---|---|
| `minilm-l12-grape-route` | Traffic controller — routes input to correct expert | [Explore →](https://huggingface.co/jrodriiguezg/minilm-l12-grape-route) |

## Training

- **Base model:** `Salesforce/codet5-small`
- **Fine-tuning platform:** Google Colab
- **Custom datasets:** Each expert trained on domain-specific custom datasets
- **Notable dataset:** [`malbec-nl2docker-es`](https://huggingface.co/datasets/jrodriiguezg/malbec-nl2docker-es) — public NL-to-Docker dataset in Spanish

## Key Design Decisions

- **Zero-Modification Policy (Pinot):** The search expert ignores install/delete requests entirely, focusing purely on information retrieval — a deliberate safety boundary.
- **Lightweight inference:** `codet5-small` enables running all experts on constrained hardware without GPU requirements.
- **Modular routing:** New expert domains can be added without retraining the existing models.

## Support

If you find this project useful:  
☕ [Support on Ko-fi](https://ko-fi.com/jrodriiguezg)

## Author

**Juan Raul Rodriguez Gil**  
[jrodriiguezg.link](https://jrodriiguezg.link)

---

© 2026 Juan Raul Rodriguez Gil