fix kokoro replacement, add 4096 batch size option

Author: LostRuins

koboldcpp.py

@@ -4501,8 +4501,8 @@ def hide_tooltip(event):
 
     tabcontent = {}
     # slider data
-    blasbatchsize_values = ["-1", "16", "32", "64", "128", "256", "512", "1024", "2048"]
-    blasbatchsize_text = ["Don't Batch BLAS", "16","32","64","128","256","512","1024","2048"]
+    blasbatchsize_values = ["-1","16","32","64","128","256","512","1024","2048","4096"]
+    blasbatchsize_text = ["Don't Batch BLAS","16","32","64","128","256","512","1024","2048","4096"]
     contextsize_text = ["256", "512", "1024", "2048", "3072", "4096", "6144", "8192", "10240", "12288", "14336", "16384", "20480", "24576", "28672", "32768", "40960", "49152", "57344", "65536", "81920", "98304", "114688", "131072"]
     antirunopts = [opt.replace("Use ", "") for lib, opt in lib_option_pairs if opt not in runopts]
     quantkv_text = ["F16 (Off)","8-Bit","4-Bit"]
@@ -7572,7 +7572,7 @@ def range_checker(arg: str):
     advparser.add_argument("--version", help="Prints version and exits.", action='store_true')
     advparser.add_argument("--analyze", metavar=('[filename]'), help="Reads the metadata, weight types and tensor names in any GGUF file.", default="")
     advparser.add_argument("--maingpu", help="Only used in a multi-gpu setup. Sets the index of the main GPU that will be used.",metavar=('[Device ID]'), type=int, default=-1)
-    advparser.add_argument("--blasbatchsize", help="Sets the batch size used in BLAS processing (default 512). Setting it to -1 disables BLAS mode, but keeps other benefits like GPU offload.", type=int,choices=[-1,16,32,64,128,256,512,1024,2048], default=512)
+    advparser.add_argument("--blasbatchsize", help="Sets the batch size used in BLAS processing (default 512). Setting it to -1 disables BLAS mode, but keeps other benefits like GPU offload.", type=int,choices=[-1,16,32,64,128,256,512,1024,2048,4096], default=512)
     advparser.add_argument("--blasthreads", help="Use a different number of threads during BLAS if specified. Otherwise, has the same value as --threads",metavar=('[threads]'), type=int, default=0)
     advparser.add_argument("--lora", help="GGUF models only, applies a lora file on top of model.", metavar=('[lora_filename]'), nargs='+')
     advparser.add_argument("--loramult", metavar=('[amount]'), help="Multiplier for the Text LORA model to be applied.", type=float, default=1.0)

otherarch/ttscpp/src/kokoro_model.cpp

@@ -1424,6 +1424,7 @@ int kokoro_runner::generate(std::string prompt, struct tts_response * response,
     // We preserve the other punctuation for cleaner chunking pre-tokenization
     prompt = replace_any(prompt, ";:", "--");
     prompt = replace_any(prompt, "\n", ". ");
+	kokoro_str_replace_all(prompt,"’","'");
 	kokoro_str_replace_all(prompt," - "," -- ");
 	kokoro_str_replace_all(prompt,"he's ","he is ");
 	kokoro_str_replace_all(prompt,"'s ","s ");