prism-ml/Bonsai-8B-mlx-1bit
Quantization vs prism-ml/Bonsai-8B-unpacked. prism-ml/Bonsai-8B-mlx-1bit is a quantized version of the other (U32/F16).
2 derivatives checked
- VertexAGI/prism-roleplay-1.5-small-fast: Same weights, repacked
- inferencerlabs/Bonsai-8B-MLX-Q2: Partial fine-tune
3 other repos hold byte-for-byte identical weight files.