@@ -257,7 +257,7 @@ def set_input_parameters(self, parameters):
257
257
def unload (self ):
258
258
print ("Attemping to unload library" )
259
259
self .process .terminate ()
260
-
260
+
261
261
262
262
def _load (self , save_model : bool , initial_load : bool ) -> None :
263
263
self .tokenizer = self ._get_tokenizer ("gpt2" )
@@ -268,8 +268,8 @@ def _load(self, save_model: bool, initial_load: bool) -> None:
268
268
unbantokens = False , bantokens = None , usemirostat = None , forceversion = 0 , nommap = self .kcpp_nommap ,
269
269
usemlock = False , noavx2 = self .kcpp_noavx2 , debugmode = self .kcpp_debugmode , skiplauncher = True , hordeconfig = None , noblas = self .kcpp_noblas ,
270
270
useclblast = self .kcpp_useclblast , usecublas = self .kcpp_usecublas , gpulayers = self .kcpp_gpulayers , tensor_split = self .kcpp_tensor_split , config = None ,
271
- onready = '' , multiuser = False , foreground = False , preloadstory = None , noshift = False , remotetunnel = False )
272
-
271
+ onready = '' , multiuser = False , foreground = False , preloadstory = None , noshift = False , remotetunnel = False , ssl = False )
272
+
273
273
274
274
#koboldcpp.main(kcppargs,False) #initialize library without enabling Lite http server
275
275
(self .output_queue , self .input_queue , self .process ) = koboldcpp .start_in_seperate_process (kcppargs )
@@ -301,17 +301,17 @@ def _raw_generate(
301
301
# Store context in memory to use it for comparison with generated content
302
302
utils .koboldai_vars .lastctx = decoded_prompt
303
303
304
- self .input_queue .put ({'command' : 'generate' , 'data' : [(decoded_prompt ,max_new ,utils .koboldai_vars .max_length ,
304
+ self .input_queue .put ({'command' : 'generate' , 'data' : [(decoded_prompt ,"" , max_new ,utils .koboldai_vars .max_length ,
305
305
gen_settings .temp ,int (gen_settings .top_k ),gen_settings .top_a ,gen_settings .top_p ,
306
306
gen_settings .typical ,gen_settings .tfs ,gen_settings .rep_pen ,gen_settings .rep_pen_range ),
307
307
{"sampler_order" : gen_settings .sampler_order , "use_default_badwordsids" : utils .koboldai_vars .use_default_badwordsids }
308
308
]})
309
-
310
- #genresult = koboldcpp.generate(decoded_prompt,max_new,utils.koboldai_vars.max_length,
309
+
310
+ #genresult = koboldcpp.generate(decoded_prompt,"", max_new,utils.koboldai_vars.max_length,
311
311
#gen_settings.temp,int(gen_settings.top_k),gen_settings.top_a,gen_settings.top_p,
312
312
#gen_settings.typical,gen_settings.tfs,gen_settings.rep_pen,gen_settings.rep_pen_range,
313
313
#sampler_order=gen_settings.sampler_order,use_default_badwordsids=utils.koboldai_vars.use_default_badwordsids)
314
-
314
+
315
315
genresult = []
316
316
while True :
317
317
data = self .output_queue .get ()
0 commit comments