prism-ml/Ternary-Bonsai-8B-mlx-2bit

Quantization vs prism-ml/Ternary-Bonsai-8B-unpacked. prism-ml/Ternary-Bonsai-8B-mlx-2bit is a quantized version of the other (U32/F16).

2 other repos hold byte-for-byte identical weight files.

On the Hub · Downloads on Model Pulse