clef-oQ6e
This model was quantized using oQ (oMLX v0.7.1rc1) mixed-precision quantization.
The effective average is 6.856 bits/weight for the complete checkpoint.
Quantization and Bit Distribution
| Storage format | Logical weights | Share of total | Tensor storage | Effective bits/weight |
|---|---|---|---|---|
| Affine 6-bit, group 64 | 24.812B | 90.274% | 18.775 GiB | 6.50 |
| Affine 8-bit, group 64 | 2.082B | 7.574% | 2.060 GiB | 8.50 |
| BF16 | 0.591B | 2.152% | 1.102 GiB | 16.00 |
| Total | 27.485B | 100% | 21.937 GiB | 6.856 |
Storage Breakdown
| Component | Logical weights | Storage |
|---|---|---|
| Language model | 26.896B | 22.377 GB / 20.840 GiB |
| Vision | 0.461B | 0.921 GB / 0.858 GiB |
| Joint schema head | 0.128B | 0.256 GB / 0.239 GiB |
| Total | 27.485B | 23.554 GB / 21.937 GiB |
- Downloads last month
- 20
Model size
27B params
Tensor type
U32
·
BF16 ·
Hardware compatibility
Log In to add your hardware
6-bit
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support