vrr/koboldcpp
2
0
Fork 0
mirror of https://github.com/LostRuins/koboldcpp.git synced 2025-09-10 09:04:36 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

various fixes

Browse source
This commit is contained in:
Concedo 2025-01-18 23:52:39 +08:00
parent e90866fd46
commit fa7e661133
14 changed files with 51 additions and 32 deletions
Download patch file Download diff file Expand all files Collapse all files

6
koboldcpp.py
View file

@ -60,7 +60,7 @@ maxhordelen = 400
modelbusy = threading.Lock()
requestsinqueue = 0
defaultport = 5001
KcppVersion = "1.82"
KcppVersion = "1.82.1"
showdebug = True
guimode = False
showsamplerwarning = True
@ -291,6 +291,7 @@ class tts_load_model_inputs(ctypes.Structure):
("cublas_info", ctypes.c_int),
("vulkan_info", ctypes.c_char_p),
("gpulayers", ctypes.c_int),
("flash_attention", ctypes.c_bool),
("debugmode", ctypes.c_int)]
class tts_generation_inputs(ctypes.Structure):
@ -1329,7 +1330,7 @@ def whisper_generate(genparams):
inputs.prompt = prompt.encode("UTF-8")
inputs.audio_data = audio_data.encode("UTF-8")
inputs.quiet = is_quiet
lc = genparams.get("langcode", "auto")
lc = genparams.get("langcode", genparams.get("language", "auto"))
lc = lc.strip().lower() if (lc and lc.strip().lower()!="") else "auto"
inputs.langcode = lc.encode("UTF-8")
inputs.suppress_non_speech = genparams.get("suppress_non_speech", False)
@ -1347,6 +1348,7 @@ def tts_load_model(ttc_model_filename,cts_model_filename):
inputs.ttc_model_filename = ttc_model_filename.encode("UTF-8")
inputs.cts_model_filename = cts_model_filename.encode("UTF-8")
inputs.gpulayers = (999 if args.ttsgpu else 0)
inputs.flash_attention = args.flashattention
thds = args.threads
if args.ttsthreads and args.ttsthreads > 0:
ttst = int(args.ttsthreads)