Qwen3.5-9B oQe MTP
Collection
8 items • Updated
Qwen3.5-9B-oQ4e-mtp
Updated 2026-10-25: Re-quantized with oMLX v0.7.3.dev2. Please re-download if you have an earlier copy.
This model was quantized using oQ (oMLX v0.7.3.dev2) mixed-precision quantization.
The effective average is 5.164 bits/weight for the complete checkpoint.
Quantization and Bit Distribution
| Storage format | Logical weights | Share of total | Tensor storage | Effective bits/weight |
|---|---|---|---|---|
| Affine 4-bit, group 64 | 8.499B | 88.044% | 4.452 GiB | 4.50 |
| Affine 5-bit, group 64 | 0.563B | 5.830% | 0.360 GiB | 5.50 |
| Affine 6-bit, group 64 | 0.101B | 1.043% | 0.076 GiB | 6.50 |
| BF16 | 0.491B | 5.083% | 0.914 GiB | 16.00 |
| Total | 9.653B | 100% | 5.803 GiB | 5.164 |
Storage Breakdown
| Component | Logical weights | Storage |
|---|---|---|
| Language model | 8.954B | 5.134 GB / 4.781 GiB |
| Vision | 0.456B | 0.912 GB / 0.849 GiB |
| MTP head | 0.243B | 0.185 GB / 0.172 GiB |
| Total | 9.653B | 6.231 GB / 5.803 GiB |
- Downloads last month
- 129
Model size
10B params
Tensor type
U32
·
BF16 ·
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