Qwen3-Coder oMLX
Collection
4 items • Updated
Qwen3-Coder-REAP-25B-A3B-oQ4e
This model was quantized using oQ (oMLX v0.5.3) mixed-precision quantization.
Quantization details
- Model type: qwen3_moe
- Bits: 4
- Group size: 64
- Format: MLX safetensors
- Downloads last month
- 177
Model size
4B params
Tensor type
BF16
·
U32 ·
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