| ▲ | Tepix 5 hours ago | |
DS4Flash has 284B weights. 64GB? No go. | ||
| ▲ | zozbot234 4 hours ago | parent [-] | |
The native weights are 4-bit for the sparse experts, and they quantize to ~80GB with limited degradation in real-world performance. That's a viable target for 64GB with SSD streaming, though it will be slower than keeping the whole thing in RAM (especially on a M2-class machine with its slower storage). | ||