Ornith-1.5-122B oQe MTP
Collection
8 items • Updated
Ornith-1.5-122B-oQ8e-fp16-mtp
This model was quantized using oQ (oMLX v0.7.1) mixed-precision quantization.
The effective average is 8.581 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 | 3.599B | 2.877% | 3.456 GiB | 8.25 |
| Affine 8-bit, group 64 | 120.977B | 96.715% | 119.710 GiB | 8.50 |
| FP16 | 0.060B | 0.048% | 0.111 GiB | 16.00 |
| FP32 | 0.451B | 0.361% | 1.681 GiB | 32.00 |
| Total | 125.086B | 100% | 124.959 GiB | 8.581 |
Storage Breakdown
| Component | Logical weights | Storage |
|---|---|---|
| Language model | 122.112B | 129.669 GB / 120.763 GiB |
| Vision | 0.451B | 1.805 GB / 1.681 GiB |
| MTP head | 2.524B | 2.700 GB / 2.514 GiB |
| Total | 125.086B | 134.173 GB / 124.959 GiB |
- Downloads last month
- 111
Model size
125B params
Tensor type
U32
·
F16 ·
F32 ·
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
Model tree for scottlowry/Ornith-1.5-122B-oQ8e-fp16-mtp
Base model
ornith-ai/Ornith-1.5-122B