| ▲ | abraxas 4 hours ago |
| I'm not following the local mdoel scene too closely but this seems quite amazing. Is this able to be run on Apple silicon too? |
|
| ▲ | Havoc 3 hours ago | parent | next [-] |
| Their first 27B bonsai was able to run on an iphone. |
|
| ▲ | kamranjon 4 hours ago | parent | prev [-] |
| "Ternary Bonsai 2 27B reaches up to 143 tokens/second on NVIDIA GeForce RTX 5090 and 46.8 tokens/second on M5 Max. On an RTX 4090, Ternary Bonsai 2 27B consumes just 0.714 mWh/token, making it 40% more energy-efficient than an 8B model running in full-precision." |
| |
| ▲ | pizza234 3 hours ago | parent | next [-] | | Their mention of the 5090 is bit odd, since on 32 GB GPUs, Q6 fits while having better quality. Very interesting model for 16 GB GPUs though! | | |
| ▲ | sisve 3 hours ago | parent [-] | | They mention 5090 with regards to speed, Q6 will not have that speed? And speed matters a lot for many use cases | | |
| ▲ | selectodude 2 hours ago | parent [-] | | 150 tokens per second on a ternary model implies that it’s GPU bound, I’d bet a Q6 model is even faster because it’s existed longer and seen more optimization. You’d have to be insane to not run an NVFP4 quant over a ternary quant on Blackwell if they both fit. |
|
| |
| ▲ | azatom 3 hours ago | parent | prev [-] | | it is like "my fridge is 2mkm (millikilometer) from my desk"
m=0.001 h=3600 it should be just Ws or just J | | |
|
