Can I run Muse-Glimmer-30B on an RTX 4060?
No
Even its smallest file (4-bit (estimate), 16.7 GB) is more than an RTX 4060 can hold (7.20 GB usable): it needs 3 of them.
Files and context on RTX 4060
Each file of Muse-Glimmer-30B (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) |
|---|---|---|---|---|
| As published | 59.6 GB | Doesn't fit | — | — |
| Q4_K_XL | 19.7 GB | Doesn't fit | — | — |
| Q4_K_M | 16.8 GB | Doesn't fit | — | — |
| BF16 | 55.7 GB | Doesn't fit | — | — |
| UD-Q8_K_XL | 32.3 GB | Doesn't fit | — | — |
| Q8_0 | 29.6 GB | Doesn't fit | — | — |
| UD-Q6_K_XL | 26.3 GB | Doesn't fit | — | — |
| UD-Q5_K_XL | 21.8 GB | Doesn't fit | — | — |
| UD-Q5_K_L | 19.8 GB | Doesn't fit | — | — |
| UD-Q5_K_M | 19.2 GB | Doesn't fit | — | — |
| UD-Q4_K_XL | 15.9 GB | Doesn't fit | — | — |
| UD-IQ3_M | 14.1 GB | Doesn't fit | — | — |
| UD-Q3_K_XL | 13.4 GB | Doesn't fit | — | — |
| UD-IQ3_XXS | 13.1 GB | Doesn't fit | — | — |
| UD-Q2_K_XL | 12.4 GB | Doesn't fit | — | — |
| UD-IQ2_M | 12.3 GB | Doesn't fit | — | — |
| UD-IQ2_XS | 11.5 GB | Doesn't fit | — | — |
| UD-IQ2_XXS | 10.7 GB | Doesn't fit | — | — |