| --- |
| language: en |
| tags: |
| - sentence-transformers |
| - intent-classification |
| - feature-extraction |
| license: mit |
| datasets: |
| - custom |
| metrics: |
| - cosine-similarity |
| --- |
| |
| # Intent Classification Model |
|
|
| This is a fine-tuned SentenceTransformer model for intent classification. It was trained on custom intent data including navigation, media controls, library management, and protocol activation commands. |
|
|
| ## Usage |
|
|
| ```python |
| from sentence_transformers import SentenceTransformer |
| model = SentenceTransformer('drithh/intent-classifier') |
| embeddings = model.encode("go to London") |
| ``` |
|
|
| ## Supported Intents |
|
|
| - **Navigation**: go to LOCATION, navigate to LOCATION |
| - **Atlas**: open atlas, launch atlas |
| - **Map Controls**: select LOCATION, show boundaries, hide boundaries |
| - **Library**: open library, close library, go to video NUMBER |
| - **Media Controls**: play video, pause video, rewind, forward |
| - **News**: show news LOCATION, hide news |
| - **Protocols**: activate PROTOCOL, deactivate PROTOCOL |
|
|
| ## Model Details |
|
|
| - **Base Model**: sentence-transformers/paraphrase-MiniLM-L3-v2 |
| - **Fine-tuning**: Cosine similarity loss |
| - **Embedding Dimensions**: 384 |
| - **Training Data**: 139,128 training pairs |
|
|
