some cleanup before starting on TTS

Author: LostRuins

Makefile

@@ -664,11 +664,12 @@ clean:
 # useful tools
 main: examples/main/main.cpp common/json-schema-to-grammar.cpp common/arg.cpp build-info.h ggml.o ggml-cpu.o llama.o console.o llavaclip_default.o llava.o ggml-backend_default.o ggml-backend-reg_default.o $(OBJS_FULL) $(OBJS)
 	$(CXX) $(CXXFLAGS) $(filter-out %.h,$^) -o $@ $(LDFLAGS)
-	@echo '====  Run ./main -h for help.  ===='
 sdmain: otherarch/sdcpp/util.cpp otherarch/sdcpp/main.cpp otherarch/sdcpp/stable-diffusion.cpp otherarch/sdcpp/upscaler.cpp otherarch/sdcpp/model.cpp otherarch/sdcpp/thirdparty/zip.c build-info.h ggml.o ggml-cpu.o llama.o console.o ggml-backend_default.o ggml-backend-reg_default.o $(OBJS_FULL) $(OBJS)
 	$(CXX) $(CXXFLAGS) $(filter-out %.h,$^) -o $@ $(LDFLAGS)
 whispermain: otherarch/whispercpp/main.cpp otherarch/whispercpp/whisper.cpp build-info.h ggml.o ggml-cpu.o llama.o console.o ggml-backend_default.o ggml-backend-reg_default.o $(OBJS_FULL) $(OBJS)
 	$(CXX) $(CXXFLAGS) $(filter-out %.h,$^) -o $@ $(LDFLAGS)
+ttsmain: examples/tts/tts.cpp common/json-schema-to-grammar.cpp common/arg.cpp build-info.h ggml.o ggml-cpu.o llama.o console.o llavaclip_default.o llava.o ggml-backend_default.o ggml-backend-reg_default.o $(OBJS_FULL) $(OBJS)
+	$(CXX) $(CXXFLAGS) $(filter-out %.h,$^) -o $@ $(LDFLAGS)
 gguf-split: examples/gguf-split/gguf-split.cpp ggml.o ggml-cpu.o llama.o build-info.h llavaclip_default.o llava.o ggml-backend_default.o ggml-backend-reg_default.o $(OBJS_FULL) $(OBJS)
 	$(CXX) $(CXXFLAGS) $(filter-out %.h,$^) -o $@ $(LDFLAGS)
 

README.md

@@ -167,15 +167,12 @@ when you can't use the precompiled binary directly, we provide an automated buil
 - If you wish, after building the koboldcpp libraries with `make`, you can rebuild the exe yourself with pyinstaller by using `make_pyinstaller.bat`
 - API documentation available at `/api` (e.g. `http://localhost:5001/api`) and https://lite.koboldai.net/koboldcpp_api. An OpenAI compatible API is also provided at `/v1` route (e.g. `http://localhost:5001/v1`).
 - **All up-to-date GGUF models are supported**, and KoboldCpp also includes backward compatibility for older versions/legacy GGML `.bin` models, though some newer features might be unavailable.
-- An incomplete list of models and architectures is listed, but there are *many hundreds of other GGUF models*. In general, if it's GGUF, it should work.
-  - Llama / Llama2 / Llama3 / Alpaca / GPT4All / Vicuna / Koala / Pygmalion / Metharme / WizardLM
-  - Mistral / Mixtral / Miqu
-  - Qwen / Qwen2 / Yi
-  - Gemma / Gemma2
-  - GPT-2 / Cerebras
-  - Phi-2 / Phi-3
-  - GPT-NeoX / Pythia / StableLM / Dolly / RedPajama
-  - GPT-J / RWKV4 / MPT / Falcon / Starcoder / Deepseek and many more
-  - [Stable Diffusion 1.5 and SDXL safetensor models](https://github.com/LostRuins/koboldcpp/wiki#can-i-generate-images-with-koboldcpp)
-  - [LLaVA based Vision models and multimodal projectors (mmproj)](https://github.com/LostRuins/koboldcpp/wiki#what-is-llava-and-mmproj)
-  - [Whisper models for Speech-To-Text](https://huggingface.co/koboldcpp/whisper/tree/main)
+- An incomplete list of architectures is listed, but there are *many hundreds of other GGUF models*. In general, if it's GGUF, it should work.
+- Llama / Llama2 / Llama3 / Alpaca / GPT4All / Vicuna / Koala / Pygmalion / Metharme / WizardLM / Mistral / Mixtral / Miqu / Qwen / Qwen2 / Yi / Gemma / Gemma2 / GPT-2 / Cerebras / Phi-2 / Phi-3 / GPT-NeoX / Pythia / StableLM / Dolly / RedPajama / GPT-J / RWKV4 / MPT / Falcon / Starcoder / Deepseek and many, **many** more.
+
+# Where can I download AI model files?
+- The best place to get GGUF text models is huggingface. For image models, CivitAI has a good selection. Here are some to get started.
+  - Text Generation: [BookAdventures 8B](https://huggingface.co/KoboldAI/Llama-3.1-8B-BookAdventures-GGUF/resolve/main/Llama-3.1-8B-BookAdventures.Q4_K_S.gguf) or [Tiefighter 13B](https://huggingface.co/KoboldAI/LLaMA2-13B-Tiefighter-GGUF/resolve/main/LLaMA2-13B-Tiefighter.Q4_K_S.gguf) (larger model).
+  - Image Generation: [Anything v3](https://huggingface.co/admruul/anything-v3.0/resolve/main/Anything-V3.0-pruned-fp16.safetensors) or [Deliberate V2](https://huggingface.co/Yntec/Deliberate2/resolve/main/Deliberate_v2.safetensors) or [Dreamshaper SDXL](https://huggingface.co/Lykon/dreamshaper-xl-v2-turbo/resolve/main/DreamShaperXL_Turbo_v2_1.safetensors)
+  - Image Recognition MMproj: [Pick the correct one for your model architecture here](https://huggingface.co/koboldcpp/mmproj/tree/main)
+  - Speech Recognition: [Whisper models for Speech-To-Text](https://huggingface.co/koboldcpp/whisper/tree/main)

src/llama-quant.cpp

@@ -397,7 +397,10 @@ static ggml_type llama_tensor_get_type(quantize_state_impl & qs, ggml_type new_t
             case GGML_TYPE_Q4_K:   new_type = GGML_TYPE_Q5_0;   break;
             case GGML_TYPE_Q5_K:   new_type = GGML_TYPE_Q5_1;   break;
             case GGML_TYPE_Q6_K:   new_type = GGML_TYPE_Q8_0;   break;
-            default: throw std::runtime_error("\nUnsupported tensor size encountered\n");
+            default:
+                printf("\nUnsupported tensor size encountered! Will use %s for %s\n",ggml_type_name(tensor->type),name.c_str()) ;
+                new_type = tensor->type;
+            break;
         }
         if (tensor->ne[0] % ggml_blck_size(new_type) != 0) {
             new_type = GGML_TYPE_F16;
@@ -629,7 +632,7 @@ static void llama_model_quantize_impl(const std::string & fname_inp, const std::
         if (llama_model_has_encoder(&model)) {
             n_attn_layer *= 3;
         }
-        GGML_ASSERT((qs.n_attention_wv == n_attn_layer) && "n_attention_wv is unexpected");
+        GGML_ASSERT_CONTINUE((qs.n_attention_wv == n_attn_layer) && "n_attention_wv is unexpected");
     }
 
     size_t total_size_org = 0;