| title: PredictLM Playground | |
| emoji: 🟢 | |
| colorFrom: green | |
| colorTo: indigo | |
| sdk: gradio | |
| sdk_version: "5.4.0" | |
| python_version: "3.11" | |
| app_file: app.py | |
| pinned: true | |
| license: apache-2.0 | |
| models: | |
| - zerooneresearch/predictlm-mini-13m | |
| short_description: Try the 13M tabular foundation model on your own CSV. | |
| # PredictLM Playground | |
| Upload a CSV, pick a target column, get predictions from **predictlm-mini-13m** — a 13M-parameter open-weight tabular foundation model (Apache-2.0). | |
| This Space runs **single-model mode** (Mini only, no test-time training) for snappy responses. The full Duo + TTT recipe — **0.751 classification accuracy / 0.609 regression R²** on a locked 25-dataset OpenML benchmark — runs in one Python call: | |
| ```python | |
| from predictlm import PredictLM | |
| model = PredictLM.from_pretrained("zerooneresearch/predictlm-mini-13m") | |
| preds = model.fit(X_train, y_train).predict(X_test) | |
| ``` | |
| - **Model card**: [zerooneresearch/predictlm-mini-13m](https://huggingface.co/zerooneresearch/predictlm-mini-13m) | |
| - **PyPI**: [`pip install predictlm`](https://pypi.org/project/predictlm/) | |
| - **Source (MCP server)**: [github.com/matej-01RAI/predictlm-mcp](https://github.com/matej-01RAI/predictlm-mcp) | |