Hy4-flash MLX 3.8-bit

A 103.65 GB quant of Hy4-flash from Nace.AI, 3.8 bits per weight.

Hy4-flash quantization quality across 37 measured variants

37 measured Hy4-flash variants on the same 11 texts, 5,621 scored tokens. Chart sizes use text-weight payload, including required Engram tables, excluding vision/MTP. Complete weight-file sizes and exact scores are in the comparison data.

Download

hf download nace-ai/Hy4-flash-MLX-3.8bit --local-dir Hy4-flash-MLX-3.8bit

This pack needs a runtime with packed-expert and GGUF Engram-table support.

License

HY Community License.

Downloads last month
134
Safetensors
Model size
219B params
Tensor type
U32
·
F32
·
BF16
·
I32
·
MLX
Hardware compatibility
Log In to add your hardware

4-bit

Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for nace-ai/Hy4-flash-MLX-3.8bit

Quantized
(48)
this model

Collection including nace-ai/Hy4-flash-MLX-3.8bit