Spaces:
Sleeping
Sleeping
| title: Model Compare | |
| emoji: 👀 | |
| colorFrom: red | |
| colorTo: indigo | |
| sdk: gradio | |
| sdk_version: 5.44.0 | |
| app_file: app.py | |
| pinned: false | |
| license: cc-by-sa-4.0 | |
| short_description: Compare our models on *your* prompts! | |
| # MediaBiasGroup — Model Comparator (Gradio Space) | |
| This Space discovers models under the `mediabiasgroup` organization and lets users: | |
| - select a **task** (e.g., `text-classification`), | |
| - compare **multiple models** on the **same input text**, | |
| - inspect any **reported metrics** (from `model-index` in model cards). | |
| ### Notes | |
| - Inference uses `transformers.pipeline` and loads models lazily. | |
| - If a model doesn’t have `id2label/label2id`, the app still shows raw labels. | |
| - The metrics table appears only if the model card includes a `model-index`. | |
| ### Extending | |
| - Add more task handlers in `get_pipeline(...)` if you release models for other tasks. | |
| - If you want to restrict the list to **Collections only**, you can replace discovery with a static list or build a small scraper for your collections page. | |