I tried loading llama-2-coder-7b.gguf.q5_k_m.bin with lmstudio chat but the response I get adds adds a bunch of undesired characters inbetween.

#1
by blacksilver1234 - opened

The issue is that it prints response like this:
How▁are▁you▁doing?<0X0A><0x0A>def▁say_hi(name):<0x0A>print('Hello▁'▁+▁name)<0x0A>say_hi('John')

How to get rid of <0x0A>. I tried a bunch of preset at it all have the same issue.

Sign up or log in to comment