vbatts/llama.cpp
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

fixed stop sequence crash

Browse source
This commit is contained in:
Concedo 2023-05-02 14:56:50 +08:00
parent 94827172e0
commit 6f702f2700
1 changed files with 6 additions and 7 deletions
Download patch file Download diff file Expand all files Collapse all files

13
koboldcpp.py
View file

@ -161,13 +161,12 @@ def generate(prompt,max_length=20, max_context_length=512,temperature=0.8,top_k=
inputs.tfs = tfs
inputs.rep_pen = rep_pen
inputs.rep_pen_range = rep_pen_range
inputs.seed = seed
if stop_sequence:
for n in range(0,stop_token_max):
if n >= len(stop_sequence):
inputs.stop_sequence[n] = "".encode("UTF-8")
else:
inputs.stop_sequence[n] = stop_sequence[n].encode("UTF-8")
inputs.seed = seed
for n in range(0,stop_token_max):
if not stop_sequence or n >= len(stop_sequence):
inputs.stop_sequence[n] = "".encode("UTF-8")
else:
inputs.stop_sequence[n] = stop_sequence[n].encode("UTF-8")
ret = handle.generate(inputs,outputs)
if(ret.status==1):
return ret.text.decode("UTF-8","ignore")