Can I run Winnow-12B on an RTX 4060?
No
Even its smallest file (Winnow-12B-NVFP4, 8.16 GB) is more than an RTX 4060 can hold (7.20 GB usable): it needs 2 of them.
Files and context on RTX 4060
Each file of Winnow-12B (and of GGUF versions the site has analyzed), and how much context fits next to it, with the context cache in 16-bit or 8-bit.
| File | Size | On RTX 4060 | Context (16-bit) | Context (8-bit) |
|---|---|---|---|---|
| Winnow-12B-BF16 | 23.8 GB | Doesn't fit | — | — |
| Q8_0 | 12.7 GB | Doesn't fit | — | — |
| Winnow-12B-NVFP4 | 8.16 GB | Doesn't fit | — | — |