clef-oQ2e
This model was quantized using oQ (oMLX v0.7.1rc1) mixed-precision quantization.
The effective average is 3.358 bits/weight for the complete checkpoint.
Quantization and Bit Distribution
| Storage format | Logical weights | Share of total | Tensor storage | Effective bits/weight |
|---|---|---|---|---|
| Affine 2-bit, group 64 | 22.968B | 83.567% | 6.685 GiB | 2.50 |
| Affine 3-bit, group 64 | 0.000B | 0.001% | 0.000 GiB | 3.50 |
| Affine 5-bit, group 64 | 2.648B | 9.636% | 1.696 GiB | 5.50 |
| Affine 6-bit, group 64 | 0.005B | 0.019% | 0.004 GiB | 6.50 |
| Affine 8-bit, group 64 | 1.271B | 4.626% | 1.258 GiB | 8.50 |
| BF16 | 0.591B | 2.152% | 1.102 GiB | 16.00 |
| Total | 27.485B | 100% | 10.744 GiB | 3.358 |
Storage Breakdown
| Component | Logical weights | Storage |
|---|---|---|
| Language model | 26.896B | 10.359 GB / 9.647 GiB |
| Vision | 0.461B | 0.921 GB / 0.858 GiB |
| Joint schema head | 0.128B | 0.256 GB / 0.239 GiB |
| Total | 27.485B | 11.536 GB / 10.744 GiB |
- Downloads last month
- 22
Model size
27B params
Tensor type
U32
·
BF16 ·
Hardware compatibility
Log In to add your hardware
2-bit
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support