flint-oQ8e-fp16
This model was quantized using oQ (oMLX v0.7.3rc1) mixed-precision quantization.
The effective average is 8.500 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 128 | 0.754B | 1.896% | 0.724 GiB | 8.25 |
| Affine 8-bit, group 64 | 39.006B | 98.042% | 38.598 GiB | 8.50 |
| FP16 | 0.025B | 0.063% | 0.046 GiB | 16.00 |
| Total | 39.785B | 100% | 39.369 GiB | 8.500 |
Storage Breakdown
| Component | Logical weights | Storage |
|---|---|---|
| Language model | 39.785B | 42.272 GB / 39.369 GiB |
| Total | 39.785B | 42.272 GB / 39.369 GiB |
- Downloads last month
- 67
Model size
40B params
Tensor type
U32
·
F16 ·
Hardware compatibility
Log In to add your hardware
8-bit
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support