eh bro i recommend removing all quants of opencrystal branches every model.. those were beta..
#1
by
Darkknight535
- opened
This is the original one.
Darkknight535/OpenCrystal-12B-L3
Currently nobody quantized Darkknight535/OpenCrystal-12B-L3
. I recommend you request it under https://huggingface.co/mradermacher/model_requests/discussions as it seems kind of strange to remove the old version before the new one got quantized.
If I understand this and your other discussion item, quanting OpenCrystal-12B-L3 should resolve this? It's currently in the queue.
mradermacher
changed discussion status to
closed
the issue is that old models was just playing around, but Darkknight535/OpenCrystal-12B-L3 this one turned out good, those were output gebberish bla bla bla.....