diff --git a/CMakeLists.txt b/CMakeLists.txt
index dc86b6b7..952cfb77 100644
--- a/CMakeLists.txt
+++ b/CMakeLists.txt
@@ -51,6 +51,20 @@ if(SD_FLASH_ATTN)
     add_definitions(-DSD_USE_FLASH_ATTENTION)
 endif()
 
+set(SD_LIB stable-diffusion)
+
+add_library(${SD_LIB} stable-diffusion.h stable-diffusion.cpp model.h model.cpp util.h util.cpp upscaler.cpp
+             ggml_extend.hpp clip.hpp common.hpp unet.hpp tae.hpp esrgan.hpp lora.hpp denoiser.hpp rng.hpp rng_philox.hpp)
+
+if(BUILD_SHARED_LIBS)
+    message("Build shared library")
+    add_definitions(-DSD_BUILD_SHARED_LIB)
+    target_compile_definitions(${SD_LIB} PRIVATE -DSD_BUILD_DLL)
+    set(CMAKE_POSITION_INDEPENDENT_CODE ON)
+else()
+    message("Build static library")
+endif()
+
 
 set(CMAKE_POLICY_DEFAULT_CMP0077 NEW)
 # deps
@@ -58,10 +72,6 @@ add_subdirectory(ggml)
 
 add_subdirectory(thirdparty)
 
-set(SD_LIB stable-diffusion)
-
-add_library(${SD_LIB} stable-diffusion.h stable-diffusion.cpp model.h model.cpp util.h util.cpp upscaler.cpp
-             ggml_extend.hpp clip.hpp common.hpp unet.hpp tae.hpp esrgan.hpp lora.hpp denoiser.hpp rng.hpp rng_philox.hpp)
 target_link_libraries(${SD_LIB} PUBLIC ggml zip)
 target_include_directories(${SD_LIB} PUBLIC . thirdparty)
 target_compile_features(${SD_LIB} PUBLIC cxx_std_11)
diff --git a/examples/cli/main.cpp b/examples/cli/main.cpp
index d01e8635..25ed7664 100644
--- a/examples/cli/main.cpp
+++ b/examples/cli/main.cpp
@@ -460,13 +460,13 @@ void sd_log_cb(enum sd_log_level_t level, const char* log, void* data) {
         return;
     }
     if (level <= SD_LOG_INFO) {
-        fprintf(stdout, log);
+        fputs(log, stdout);
         fflush(stdout);
     } else {
-        fprintf(stderr, log);
+        fputs(log, stderr);
         fflush(stderr);
     }
-};
+}
 
 int main(int argc, const char* argv[]) {
     SDParams params;
@@ -560,6 +560,7 @@ int main(int argc, const char* argv[]) {
 
     if (results == NULL) {
         printf("generate failed\n");
+        free_sd_ctx(sd_ctx);
         return 1;
     }
 
@@ -600,6 +601,8 @@ int main(int argc, const char* argv[]) {
         free(results[i].data);
         results[i].data = NULL;
     }
+    free(results);
+    free_sd_ctx(sd_ctx);
 
     return 0;
 }
diff --git a/model.h b/model.h
index 865c9b0d..3b512223 100644
--- a/model.h
+++ b/model.h
@@ -6,6 +6,7 @@
 #include <memory>
 #include <string>
 #include <vector>
+#include <set>
 
 #include "ggml/ggml-backend.h"
 #include "ggml/ggml.h"
diff --git a/stable-diffusion.cpp b/stable-diffusion.cpp
index f67d3382..10e24585 100644
--- a/stable-diffusion.cpp
+++ b/stable-diffusion.cpp
@@ -101,6 +101,7 @@ class StableDiffusionGGML {
     }
 
     ~StableDiffusionGGML() {
+        ggml_backend_free(backend);
     }
 
     bool load_from_file(const std::string& model_path,
@@ -626,7 +627,7 @@ class StableDiffusionGGML {
 
                     // get_ancestral_step
                     float sigma_up   = std::min(sigmas[i + 1],
-                                              std::sqrt(sigmas[i + 1] * sigmas[i + 1] * (sigmas[i] * sigmas[i] - sigmas[i + 1] * sigmas[i + 1]) / (sigmas[i] * sigmas[i])));
+                                                std::sqrt(sigmas[i + 1] * sigmas[i + 1] * (sigmas[i] * sigmas[i] - sigmas[i + 1] * sigmas[i + 1]) / (sigmas[i] * sigmas[i])));
                     float sigma_down = std::sqrt(sigmas[i + 1] * sigmas[i + 1] - sigma_up * sigma_up);
 
                     // Euler method
@@ -802,7 +803,7 @@ class StableDiffusionGGML {
 
                     // get_ancestral_step
                     float sigma_up   = std::min(sigmas[i + 1],
-                                              std::sqrt(sigmas[i + 1] * sigmas[i + 1] * (sigmas[i] * sigmas[i] - sigmas[i + 1] * sigmas[i + 1]) / (sigmas[i] * sigmas[i])));
+                                                std::sqrt(sigmas[i + 1] * sigmas[i + 1] * (sigmas[i] * sigmas[i] - sigmas[i + 1] * sigmas[i + 1]) / (sigmas[i] * sigmas[i])));
                     float sigma_down = std::sqrt(sigmas[i + 1] * sigmas[i + 1] - sigma_up * sigma_up);
                     auto t_fn        = [](float sigma) -> float { return -log(sigma); };
                     auto sigma_fn    = [](float t) -> float { return exp(-t); };