3 ms·
RAM requirements? My current rule of thumb is “a byte per parameter”, but I doubt this runs in 1/9th that (~ 3GiB). Also, perf speedup?
by hedora 9d ago
RAM requirements? My current rule of thumb is “a byte per parameter”, but I doubt this runs in 1/9th that (~ 3GiB).
Also, perf speedup?
- jjcm 9d agoI'm seeing around 7.9GB of ram, 120 tokens/s on a 6000 pro blackwell.
- kennywinker 8d agoWithout context, it should be number of params * 1.76 (the “effective bits per weight”) / 8 So for this one, 27B * 1.76 / 8 = 5.94 GB For speed, far as I can tell it depends if your gpu is memory bandwidth bound or (mostly older gpus) processing bound. If it’s memory bandwidth bound, and your gpu gets 300GB/s, that’s: 300GB/s / 5.98 GB = 50.5t/s. Realistically it’s probably a bit slower, but that is your theoretical maximum.
- Dwedit 8d agoIt needs more VRAM than just the model weights. With 6GB of VRAM, I got 44/65 layers loaded into VRAM. Has anyone tested 8GB yet?
- kennywinker 8d ago“Without context” was me gesturing at that. Interesting you can’t fit the whole model tho - why is beyond my current understanding :)
- Dwedit 8d agoEven with a small context size (4096 at 64KB per token), that's like 256MB for the context. It's more than just the context that's eating up VRAM.