Text Classification
MLX
Safetensors
English
gemma4
mlx-4bit
decision-model
classification
calibration
baseten
apple-silicon
4-bit precision
Instructions to use pratikgx/bison-1-MLX-4bit with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- MLX
How to use pratikgx/bison-1-MLX-4bit with MLX:
# Download the model from the Hub pip install huggingface_hub[hf_xet] hf download pratikgx/bison-1-MLX-4bit --local-dir bison-1-MLX-4bit
- Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- LM Studio
- Atomic Chat
Download engine/__init__.py from pratikgx/bison-1-MLX-4bit: direct link, hf CLI and curl.
- Browser
- Download file 523 Bytes
-
https://huggingface.co/pratikgx/bison-1-MLX-4bit/resolve/main/engine/__init__.py
- Command line
-
hf download hf://pratikgx/bison-1-MLX-4bit/engine/__init__.py
-
curl -L -o __init__.py https://huggingface.co/pratikgx/bison-1-MLX-4bit/resolve/main/engine/__init__.py
523 Bytes
| """MLX System One engine for Baseten bison-1 and groundhog-1. | |
| Prompt rendering, slot labels, temperatures and answer assembly follow OneJev/qev | |
| (qev-labels-v2) plus Baseten's serve.patch. Attention and readout follow the | |
| checkpoint: `baseten_modeling_overrides` / `modeling_overrides` and `readout.json` | |
| (block-bidirectional full-attention, causal sliding-window, slot_lm, 256 slots). | |
| """ | |
| from .engine_mlx import DecisionEngine | |
| from .readout import load_readout_config | |
| __all__ = ["DecisionEngine", "load_readout_config"] | |