| ▲ | naraic0o 3 hours ago | |
you mean quantized? it’s only got 16B active parameters so it can be pretty fast at decode. | ||
| ▲ | swiftcoder 2 hours ago | parent [-] | |
sorry, yes, quantised | ||
| ▲ | naraic0o 3 hours ago | |
you mean quantized? it’s only got 16B active parameters so it can be pretty fast at decode. | ||
| ▲ | swiftcoder 2 hours ago | parent [-] | |
sorry, yes, quantised | ||