dicksondickson/Swift-1.5-Qwen3.8-27b-oQ8e-bf16-mtp-MLX

[!NOTE] Updated Oct 24, 2026: Re-quantized with oMLX 0.7.3.dev2 with improved oQe Quantization and Better Low-Bit oQ. Please re-download if you have an earlier copy.

This model is a quantization of: https://huggingface.co/ukisai/Swift-1.5-Qwen3.8-27b

This model was quantized using oQ (oMLX 0.7.3.dev2) imatrix, mixed-precision quantization with Better Low-Bit oQ.

Important tensors are left at bf16 which is for Apple M3 chips and later.

Apple M1/M2 use the fp16 variant.

Run the model using oMLX: https://github.com/jundot/omlx