mlx-community/Qwen3.8-27B-4bit
Quantization vs Qwen/Qwen3.8-27B. mlx-community/Qwen3.8-27B-4bit is a quantized version of the other (U32/BF16).
5 derivatives checked
- MK4-Research/MK4-Ginko-v2: Partial fine-tune
- MK4-Research/MK4-Ginko-v1: Partial fine-tune
- MK4-Research/MK4-Ginko-v1.5: Quantization
- bluehawana/Qwen3.8-27B-4bit-MTP-MLX: Same weights, repacked
- MK4-Research/LOREA-cyber-v6-pilot: Partial fine-tune
19 other repos hold byte-for-byte identical weight files.