fixed lora, now works quanted too

.gitignore

@@ -22,6 +22,8 @@ ggml/src/ggml-vulkan-shaders-noext.cpp
 ggml/src/ggml-vulkan-shaders-noext.hpp
 vulkan-shaders-gen-noext.exe
 vulkan-shaders-gen-noext
+vulkan-spv-tmp/*
+vulkan-spv-noext-tmp/*
 
 lcov-report/
 gcovr-report/

koboldcpp.py

@@ -3690,7 +3690,6 @@ def show_gui():
                     toggleflashattn(1,1,1)
                     togglectxshift(1,1,1)
                     togglehorde(1,1,1)
-                    togglesdquant(1,1,1)
                     toggletaesd(1,1,1)
                     tabbuttonaction(tabnames[curr_tab_idx])
 
@@ -4400,23 +4399,10 @@ def show_gui():
     makelabelentry(images_tab, "Image Threads:" , sd_threads_var, 6, 50,padx=290,singleline=True,tooltip="How many threads to use during image generation.\nIf left blank, uses same value as threads.")
     sd_model_var.trace("w", gui_changed_modelfile)
 
-    sdloritem1,sdloritem2,sdloritem3 = makefileentry(images_tab, "Image LoRA (Must be non-quant):", "Select SD lora file",sd_lora_var, 10, width=280, singlecol=True, filetypes=[("*.safetensors *.gguf", "*.safetensors *.gguf")],tooltiptxt="Select a .safetensors or .gguf SD LoRA model file to be loaded.")
+    makefileentry(images_tab, "Image LoRA (safetensors/gguf):", "Select SD lora file",sd_lora_var, 10, width=280, singlecol=True, filetypes=[("*.safetensors *.gguf", "*.safetensors *.gguf")],tooltiptxt="Select a .safetensors or .gguf SD LoRA model file to be loaded. Should be unquantized!")
-    sdloritem4,sdloritem5 = makelabelentry(images_tab, "Image LoRA Multiplier:" , sd_loramult_var, 12, 50,padx=290,singleline=True,tooltip="What mutiplier value to apply the SD LoRA with.")
+    makelabelentry(images_tab, "Image LoRA Multiplier:" , sd_loramult_var, 12, 50,padx=290,singleline=True,tooltip="What mutiplier value to apply the SD LoRA with.")
-    def togglesdquant(a,b,c):
+
-        if sd_quant_var.get()==1:
+    makecheckbox(images_tab, "Compress Weights (Saves Memory)", sd_quant_var, 8,tooltiptxt="Quantizes the SD model weights to save memory. May degrade quality.")
-            sdloritem1.grid_remove()
-            sdloritem2.grid_remove()
-            sdloritem3.grid_remove()
-            sdloritem4.grid_remove()
-            sdloritem5.grid_remove()
-        else:
-            if not sdloritem1.grid_info() or not sdloritem2.grid_info() or not sdloritem3.grid_info() or not sdloritem4.grid_info() or not sdloritem5.grid_info():
-                sdloritem1.grid()
-                sdloritem2.grid()
-                sdloritem3.grid()
-                sdloritem4.grid()
-                sdloritem5.grid()
-    makecheckbox(images_tab, "Compress Weights (Saves Memory)", sd_quant_var, 8,command=togglesdquant,tooltiptxt="Quantizes the SD model weights to save memory. May degrade quality.")
     sd_quant_var.trace("w", changed_gpulayers_estimate)
 
     makefileentry(images_tab, "T5-XXL File:", "Select Optional T5-XXL model file (SD3 or flux)",sd_t5xxl_var, 14, width=280, singlerow=True, filetypes=[("*.safetensors *.gguf","*.safetensors *.gguf")],tooltiptxt="Select a .safetensors t5xxl file to be loaded.")
@@ -4665,13 +4651,11 @@ def show_gui():
             args.sdclipg = sd_clipg_var.get()
         if sd_quant_var.get()==1:
             args.sdquant = True
-            args.sdlora = ""
+        if sd_lora_var.get() != "":
+            args.sdlora = sd_lora_var.get()
+            args.sdloramult = float(sd_loramult_var.get())
         else:
-            if sd_lora_var.get() != "":
+            args.sdlora = ""
-                args.sdlora = sd_lora_var.get()
-                args.sdloramult = float(sd_loramult_var.get())
-            else:
-                args.sdlora = ""
 
         if whisper_model_var.get() != "":
             args.whispermodel = whisper_model_var.get()
@@ -6531,7 +6515,7 @@ if __name__ == '__main__':
     sdparsergroupvae.add_argument("--sdvaeauto", help="Uses a built-in VAE via TAE SD, which is very fast, and fixed bad VAEs.", action='store_true')
     sdparsergrouplora = sdparsergroup.add_mutually_exclusive_group()
     sdparsergrouplora.add_argument("--sdquant", help="If specified, loads the model quantized to save memory.", action='store_true')
-    sdparsergrouplora.add_argument("--sdlora", metavar=('[filename]'), help="Specify a stable diffusion LORA safetensors model to be applied. Cannot be used with quant models.", default="")
+    sdparsergrouplora.add_argument("--sdlora", metavar=('[filename]'), help="Specify a stable diffusion LORA safetensors model to be applied.", default="")
     sdparsergroup.add_argument("--sdloramult", metavar=('[amount]'), help="Multiplier for the LORA model to be applied.", type=float, default=1.0)
     sdparsergroup.add_argument("--sdnotile", help="Disables VAE tiling, may not work for large images.", action='store_true')
 

otherarch/sdcpp/lora.hpp

@@ -3,7 +3,7 @@
 
 #include "ggml_extend.hpp"
 
-#define LORA_GRAPH_SIZE 10240
+#define LORA_GRAPH_SIZE 20480
 
 struct LoraModel : public GGMLRunner {
     enum lora_t {