added override tensor

Author: LostRuins

expose.h

@@ -62,6 +62,7 @@ struct load_model_inputs
     const int moe_experts = -1;
     const bool no_bos_token = false;
     const char * override_kv = nullptr;
+    const char * override_tensors = nullptr;
     const bool flash_attention = false;
     const float tensor_split[tensor_split_max] = {};
     const int quant_k = 0;

gpttype_adapter.cpp

@@ -2172,6 +2172,8 @@ ModelLoadResult gpttype_load_model(const load_model_inputs inputs, FileFormat in
         }
 
         std::vector<llama_model_kv_override> kvos; //ensure it keeps in scope until model is created
+        std::vector<llama_model_tensor_buft_override> tenos; //ensure it keeps in scope until model is created
+        std::vector<std::string> temp_tensor_names; //store temp tensor names to have mem references.
         if(inputs.moe_experts>0)
         {
             printf("\nOverriding number of experts to %d\n",inputs.moe_experts);
@@ -2195,13 +2197,58 @@ ModelLoadResult gpttype_load_model(const load_model_inputs inputs, FileFormat in
         {
             printf("\nAttempting to apply KV override: %s...\n",override_kv.c_str());
             bool kvo_ok = string_parse_kv_override(override_kv.c_str(),kvos);
-            LLAMA_LOG_INFO("\nKV override result: %s\n",(kvo_ok?"success":"failed"));
+            LLAMA_LOG_INFO("\nKV override parse: %s\n",(kvo_ok?"success":"failed"));
             fflush(stdout);
         }
         if(kvos.size()>0)
         {
+            kvos.emplace_back();
+            kvos.back().key[0] = 0;
             model_params.kv_overrides = kvos.data();
         }
+        //handle override tensor
+        std::string tensoroverrides = inputs.override_tensors;
+        if(tensoroverrides!="" && ggml_backend_dev_count()>1)
+        {
+            printf("Handling Override Tensors for backends: ");
+            std::map<std::string, ggml_backend_buffer_type_t> buft_list;
+            for (size_t i = 0; i < ggml_backend_dev_count(); ++i) {
+                auto *      dev  = ggml_backend_dev_get(i);
+                auto *      buft = ggml_backend_dev_buffer_type(dev);
+                if (buft) {
+                    std::string name = ggml_backend_buft_name(buft);
+                    printf("%s ", name.c_str());
+                    buft_list[name] = buft;
+                }
+            }
+            printf("\n\n");
+            for (const auto & override : string_split<std::string>(tensoroverrides, ',')) {
+                std::string::size_type pos = override.find('=');
+                if (pos == std::string::npos) {
+                    printf("\nInvalid Override Tensor: %s\n",override.c_str());
+                    continue;
+                }
+                std::string tensor_name = override.substr(0, pos);
+                std::string buffer_type = override.substr(pos + 1);
+
+                if (buft_list.find(buffer_type) == buft_list.end()) {
+                    printf("\nUnknown Buffer Type: %s\n",buffer_type.c_str());
+                    continue;
+                }
+                llama_model_tensor_buft_override nto;
+                temp_tensor_names.push_back(tensor_name);
+                nto.pattern = temp_tensor_names[temp_tensor_names.size()-1].c_str();
+                nto.buft =  buft_list.at(buffer_type);
+                tenos.push_back(nto);
+                printf("Override Tensor: %s to %s\n",tensor_name.c_str(),buffer_type.c_str());
+            }
+        }
+        if(tenos.size()>0)
+        {
+            tenos.push_back({nullptr, nullptr});
+            model_params.tensor_buft_overrides = tenos.data();
+        }
+
         llama_model * llamamodel = llama_model_load_from_file(kcpp_data->model_filename.c_str(), model_params);
 
         if(overwriteRope)

koboldcpp.py

@@ -181,6 +181,7 @@ class load_model_inputs(ctypes.Structure):
                 ("moe_experts", ctypes.c_int),
                 ("no_bos_token", ctypes.c_bool),
                 ("override_kv", ctypes.c_char_p),
+                ("override_tensors", ctypes.c_char_p),
                 ("flash_attention", ctypes.c_bool),
                 ("tensor_split", ctypes.c_float * tensor_split_max),
                 ("quant_k", ctypes.c_int),
@@ -1214,6 +1215,7 @@ def load_model(model_filename):
     inputs.moe_experts = args.moeexperts
     inputs.no_bos_token = args.nobostoken
     inputs.override_kv = args.overridekv.encode("UTF-8") if args.overridekv else "".encode("UTF-8")
+    inputs.override_tensors = args.overridetensors.encode("UTF-8") if args.overridetensors else "".encode("UTF-8")
     inputs = set_backend_props(inputs)
     ret = handle.load_model(inputs)
     return ret
@@ -3868,6 +3870,7 @@ def hide_tooltip(event):
     defaultgenamt_var = ctk.StringVar(value=str(512))
     nobostoken_var = ctk.IntVar(value=0)
     override_kv_var = ctk.StringVar(value="")
+    override_tensors_var = ctk.StringVar(value="")
 
     model_var = ctk.StringVar()
     lora_var = ctk.StringVar()
@@ -4393,8 +4396,9 @@ def togglerope(a,b,c):
     qkvslider,qkvlabel,qkvtitle = makeslider(tokens_tab, "Quantize KV Cache:", quantkv_text, quantkv_var, 0, 2, 30, set=0,tooltip="Enable quantization of KV cache.\nRequires FlashAttention for full effect, otherwise only K cache is quantized.")
     quantkv_var.trace("w", toggleflashattn)
     makecheckbox(tokens_tab, "No BOS Token", nobostoken_var, 43, tooltiptxt="Prevents BOS token from being added at the start of any prompt. Usually NOT recommended for most models.")
-    makelabelentry(tokens_tab, "MoE Experts:", moeexperts_var, row=45, padx=100, singleline=True, tooltip="Override number of MoE experts.")
-    makelabelentry(tokens_tab, "Override KV:", override_kv_var, row=47, padx=100, singleline=True, width=150, tooltip="Advanced option to override model metadata by key, same as in llama.cpp. Mainly for debugging, not intended for general use. Types: int, float, bool, str")
+    makelabelentry(tokens_tab, "MoE Experts:", moeexperts_var, row=45, padx=120, singleline=True, tooltip="Override number of MoE experts.")
+    makelabelentry(tokens_tab, "Override KV:", override_kv_var, row=47, padx=120, singleline=True, width=150, tooltip="Advanced option to override model metadata by key, same as in llama.cpp. Mainly for debugging, not intended for general use. Types: int, float, bool, str")
+    makelabelentry(tokens_tab, "Override Tensors:", override_tensors_var, row=49, padx=120, singleline=True, width=150, tooltip="Advanced option to override tensor backend selection, same as in llama.cpp.")
 
     # Model Tab
     model_tab = tabcontent["Loaded Files"]
@@ -4667,6 +4671,7 @@ def export_vars():
         args.defaultgenamt = int(defaultgenamt_var.get()) if defaultgenamt_var.get()!="" else 512
         args.nobostoken = (nobostoken_var.get()==1)
         args.overridekv = None if override_kv_var.get() == "" else override_kv_var.get()
+        args.overridetensors = None if override_tensors_var.get() == "" else override_tensors_var.get()
         args.chatcompletionsadapter = None if chatcompletionsadapter_var.get() == "" else chatcompletionsadapter_var.get()
         try:
             if kcpp_exporting_template and isinstance(args.chatcompletionsadapter, str) and args.chatcompletionsadapter!="" and os.path.exists(args.chatcompletionsadapter):
@@ -4861,6 +4866,8 @@ def import_vars(dict):
         nobostoken_var.set(dict["nobostoken"] if ("nobostoken" in dict) else 0)
         if "overridekv" in dict and dict["overridekv"]:
             override_kv_var.set(dict["overridekv"])
+        if "overridetensors" in dict and dict["overridetensors"]:
+            override_tensors_var.set(dict["overridetensors"])
 
         if "blasbatchsize" in dict and dict["blasbatchsize"]:
             blas_size_var.set(blasbatchsize_values.index(str(dict["blasbatchsize"])))
@@ -6588,6 +6595,7 @@ def range_checker(arg: str):
     advparser.add_argument("--nobostoken", help="Prevents BOS token from being added at the start of any prompt. Usually NOT recommended for most models.", action='store_true')
     advparser.add_argument("--maxrequestsize", metavar=('[size in MB]'), help="Specify a max request payload size. Any requests to the server larger than this size will be dropped. Do not change if unsure.", type=int, default=32)
     advparser.add_argument("--overridekv", metavar=('[name=type:value]'), help="Advanced option to override a metadata by key, same as in llama.cpp. Mainly for debugging, not intended for general use. Types: int, float, bool, str", default="")
+    advparser.add_argument("--overridetensors", metavar=('[tensor name pattern=buffer type]'), help="Advanced option to override tensor backend selection, same as in llama.cpp.", default="")
     compatgroup2 = parser.add_mutually_exclusive_group()
     compatgroup2.add_argument("--showgui", help="Always show the GUI instead of launching the model right away when loading settings from a .kcpps file.", action='store_true')
     compatgroup2.add_argument("--skiplauncher", help="Doesn't display or use the GUI launcher.", action='store_true')