Downloads · 30 days
43
24% of all-time downloads
Qwen3.5-4B-oQ8e-fp16-mtp
scottlowry/Qwen3.5-4B-oQ8e-fp16-mtp
Qwen3.5-4B-oQ8e-fp16-mtp is a machine learning model from scottlowry. Use it for the machine learning task on the model card, and read the license before you ship it in a product. It is set up for mlx.
[!NOTE] Updated 2026-11-11: Re-quantized with oMLX v0.7.4.dev1. Please re-download if you have an earlier copy.
All-time downloads
176
Public
Parameters
4.7B
6 GB on disk
Likes
1
Public
Repo makeup
Click a slice to open those files.
.safetensors5.9 GB · 100%
Parameter types
How the weights are stored.
U324.3B · 93%
Base models
Model card
From the Hugging Face model README
Qwen3.5-4B-oQ8e-fp16-mtp
[!NOTE] Updated 2026-11-11: Re-quantized with oMLX v0.7.4.dev1. Please re-download if you have an earlier copy.
This model was quantized using oQ (oMLX v0.7.4.dev1) mixed-precision quantization.
The effective average is 10.205 bits/weight for the complete checkpoint.
Quantization and Bit Distribution
| Storage format | Logical weights | Share of total | Tensor storage | Effective bits/weight |
|---|---|---|---|---|
| Affine 8-bit, group 64 | 4.312B | 92.541% | 4.267 GiB | 8.50 |
| FP16 | 0.014B | 0.302% | 0.026 GiB | 16.00 |
| FP32 | 0.334B | 7.157% | 1.242 GiB | 32.00 |
| Total | 4.660B | 100% | 5.536 GiB | 10.205 |
Storage Breakdown
| Component | Logical weights | Storage |
|---|---|---|
| Language model | 4.206B | 4.470 GB / 4.163 GiB |
| Vision | 0.334B | 1.334 GB / 1.242 GiB |
| MTP head | 0.121B | 0.140 GB / 0.131 GiB |
| Total | 4.660B | 5.944 GB / 5.536 GiB |
- Library
- mlx
- Type
- qwen3_5
- Created
- Jul 24, 2026
- Updated
- Nov 11, 2026