4 ms·
From this thread [0] I can assume that because, while 1.6T, it is A49B, it can run (theoretically, very slow maybe) locally on consumer hardeware, or is that wr
by anonzzzies 6mo ago
From this thread [0] I can assume that because, while 1.6T, it is A49B, it can run (theoretically, very slow maybe) locally on consumer hardeware, or is that wrong?
[0] https://news.ycombinator.com/item?id=47864835 https://news.ycombinator.com/item?id=47864835
- Quasimarion 6mo agoTheoretically with streaming, any model that fit the disk can run on consumer hardware, just terribly slow.
- alecco 6mo agoIf 5090 has 32GB, and let's say somehow a 1-bit quantization is possible and you don't need more VRAM for anything else (forget KV cache etc), it would be able to fit a 256B 1-bit model. Just to picture it in extremes how unlikely this is. And the active parameters come from the experts. For each token the model picks some experts to run the pass (usually 2 to 4, I haven't read V4's papers). It's not always the same experts. OTOH, being DeepSeek, I foresee a bunch of V4 distilled FP8 models fitting in a 5090 with tiny batches and with performance close from 75 to 85% of V4. And this might be good enough for many everyday tasks. Today is a good day for open models. Thank god for DeepSeek.
- imrebuild 6mo agoIt will be Seconds Per Token instead of Tokens Per Second.
- lostmsu 6mo ago2 s/t on to NVMes The issue is KV cache
- mark_l_watson 6mo agoThe flash version is smaller, I think around 200B parameters and is cheap to run.
- halJordan 6mo agoNo because you don't know which 49 is needed until the moment it's needed