Upload folder using huggingface_hub
Browse files
README.md
CHANGED
|
@@ -1,43 +1,3 @@
|
|
| 1 |
-
# Jarvis Intent Classifier
|
| 2 |
-
|
| 3 |
-
A lightweight CPU-friendly assistant model that detects user intent and returns a response + action command.
|
| 4 |
-
|
| 5 |
-
**Architecture:** SentenceTransformer (MiniLM-L6-v2) + Logistic Regression (sklearn)
|
| 6 |
-
**Trained on:** custom OS assistant dataset
|
| 7 |
-
|
| 8 |
-
### Example
|
| 9 |
-
|
| 10 |
-
Input:
|
| 11 |
-
```json
|
| 12 |
-
{
|
| 13 |
-
"intent": "system_restart",
|
| 14 |
-
"response": "System restart initiated...",
|
| 15 |
-
"action": "os.system('shutdown /r /t 1')"
|
| 16 |
-
}
|
| 17 |
-
|
| 18 |
-
|
| 19 |
-
---
|
| 20 |
-
|
| 21 |
-
## 💡 Explanation of Each Field
|
| 22 |
-
|
| 23 |
-
| Key | Description |
|
| 24 |
-
|-----|--------------|
|
| 25 |
-
| `language` | Language of your model (English = `en`) |
|
| 26 |
-
| `license` | License type (MIT, Apache-2.0, etc.) |
|
| 27 |
-
| `tags` | Helps in discoverability on Hugging Face |
|
| 28 |
-
| `model-type` | General category (e.g., `text-classification`) |
|
| 29 |
-
| `pipeline_tag` | Specifies the type of pipeline for inference |
|
| 30 |
-
| `base_model` | The pretrained embedder used (e.g., MiniLM-L6-v2) |
|
| 31 |
-
| `widget` | Adds demo examples to the model’s Hugging Face page |
|
| 32 |
-
|
| 33 |
-
---
|
| 34 |
-
|
| 35 |
-
## ✅ After Adding YAML
|
| 36 |
-
|
| 37 |
-
When you push the updated `README.md`, the warning:
|
| 38 |
-
|
| 39 |
-
|
| 40 |
-
|
| 41 |
---
|
| 42 |
language: en
|
| 43 |
license: mit
|
|
@@ -55,6 +15,37 @@ widget:
|
|
| 55 |
- text: "restart system"
|
| 56 |
---
|
| 57 |
|
| 58 |
-
# Jarvis Intent Classifier
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 59 |
|
| 60 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
---
|
| 2 |
language: en
|
| 3 |
license: mit
|
|
|
|
| 15 |
- text: "restart system"
|
| 16 |
---
|
| 17 |
|
| 18 |
+
# 🧠 Jarvis Intent Classifier
|
| 19 |
+
|
| 20 |
+
A lightweight CPU-friendly intent classification model that detects user commands
|
| 21 |
+
and returns both a natural response and a system action command.
|
| 22 |
+
|
| 23 |
+
This model is ideal for building local AI assistants that can interact with your operating system
|
| 24 |
+
(e.g., open apps, control volume, restart, sleep, or play music).
|
| 25 |
+
|
| 26 |
+
---
|
| 27 |
+
|
| 28 |
+
## 🏗️ Model Overview
|
| 29 |
+
|
| 30 |
+
| Component | Description |
|
| 31 |
+
|------------|--------------|
|
| 32 |
+
| **Base Model** | `sentence-transformers/all-MiniLM-L6-v2` (384-dim embeddings) |
|
| 33 |
+
| **Classifier** | `LogisticRegression` (scikit-learn) |
|
| 34 |
+
| **Dataset** | Custom OS command dataset with examples, responses, and Python actions |
|
| 35 |
+
| **Frameworks** | `sentence-transformers`, `scikit-learn` |
|
| 36 |
+
| **Runs on** | CPU (no GPU needed) |
|
| 37 |
+
| **License** | MIT |
|
| 38 |
+
|
| 39 |
+
---
|
| 40 |
+
|
| 41 |
+
## 🧩 Example Input → Output
|
| 42 |
|
| 43 |
+
**Input:**
|
| 44 |
+
```text
|
| 45 |
+
restart my computer
|
| 46 |
+
|
| 47 |
+
{
|
| 48 |
+
"intent": "system_restart",
|
| 49 |
+
"response": "System restart initiated...",
|
| 50 |
+
"action": "os.system('shutdown /r /t 1')"
|
| 51 |
+
}
|