README.md

@@ -123,7 +123,7 @@ API and command-line option may change frequently.***
     - Stable Diffusion v1.5 from https://huggingface.co/stable-diffusion-v1-5/stable-diffusion-v1-5 
 
     ```sh
-    curl -L -O https://huggingface.co/stable-diffusion-v1-5/stable-diffusion-v1-5/resolve/main/v1-5-pruned-emaonly.safetensors
+    curl -L -O https://huggingface.co/runwayml/stable-diffusion-v1-5/resolve/main/v1-5-pruned-emaonly.safetensors
     ```
 
 ### Generate an image with just one command

docs/sd.md

@@ -2,8 +2,8 @@
 
 - download original weights(.ckpt or .safetensors). For example
     - Stable Diffusion v1.4 from https://huggingface.co/CompVis/stable-diffusion-v-1-4-original
-    - Stable Diffusion v1.5 from https://huggingface.co/stable-diffusion-v1-5/stable-diffusion-v1-5
-    - Stable Diffuison v2.1 from https://huggingface.co/Manojb/stable-diffusion-2-1-base
+    - Stable Diffusion v1.5 from https://huggingface.co/runwayml/stable-diffusion-v1-5
+    - Stable Diffuison v2.1 from https://huggingface.co/stabilityai/stable-diffusion-2-1
     - Stable Diffusion 3 2B from https://huggingface.co/stabilityai/stable-diffusion-3-medium
 
 ### txt2img example
@@ -34,4 +34,4 @@ Using formats of different precisions will yield results of varying quality.
 
 <p align="center">
   <img src="../assets/img2img_output.png" width="256x">
-</p>
+</p>

src/core/ggml_extend.hpp

@@ -3309,7 +3309,7 @@ public:
 
 class Identity : public UnaryBlock {
 public:
-    ggml_tensor* forward(GGMLRunnerContext* ctx, ggml_tensor* x) override {
+    ggml_tensor* forward(GGMLRunnerContext* ctx, ggml_tensor* x) {
         return x;
     }
 };
@@ -3368,7 +3368,7 @@ public:
         force_prec_f32 = force_prec_f32_;
     }
 
-    ggml_tensor* forward(GGMLRunnerContext* ctx, ggml_tensor* x) override {
+    ggml_tensor* forward(GGMLRunnerContext* ctx, ggml_tensor* x) {
         ggml_tensor* w = params["weight"];
         ggml_tensor* b = nullptr;
         if (bias) {
@@ -3422,7 +3422,7 @@ public:
     }
 
     ggml_tensor* forward(GGMLRunnerContext* ctx,
-                         ggml_tensor* input_ids) override {
+                         ggml_tensor* input_ids) {
         // input_ids: [N, n_token]
         auto weight = params["weight"];
 
@@ -3482,11 +3482,11 @@ public:
         scale = scale_value;
     }
 
-    std::string get_desc() override {
+    std::string get_desc() {
         return "Conv2d";
     }
 
-    ggml_tensor* forward(GGMLRunnerContext* ctx, ggml_tensor* x) override {
+    ggml_tensor* forward(GGMLRunnerContext* ctx, ggml_tensor* x) {
         ggml_tensor* w = params["weight"];
         ggml_tensor* b = nullptr;
         if (bias) {
@@ -3569,11 +3569,11 @@ public:
         scale = scale_value;
     }
 
-    std::string get_desc() override {
+    std::string get_desc() {
         return "Conv2d_grouped";
     }
 
-    ggml_tensor* forward(GGMLRunnerContext* ctx, ggml_tensor* x) override {
+    ggml_tensor* forward(GGMLRunnerContext* ctx, ggml_tensor* x) {
         ggml_tensor* w = params["weight"];
         ggml_tensor* b = nullptr;
         if (bias) {
@@ -3725,7 +3725,7 @@ public:
           bias(bias),
           force_prec_f32(force_prec_f32) {}
 
-    ggml_tensor* forward(GGMLRunnerContext* ctx, ggml_tensor* x) override {
+    ggml_tensor* forward(GGMLRunnerContext* ctx, ggml_tensor* x) {
         ggml_tensor* w = params["weight"];
         ggml_tensor* b = nullptr;
         if (ctx->weight_adapter) {
@@ -3778,7 +3778,7 @@ public:
           elementwise_affine(elementwise_affine),
           bias(bias) {}
 
-    ggml_tensor* forward(GGMLRunnerContext* ctx, ggml_tensor* x) override {
+    ggml_tensor* forward(GGMLRunnerContext* ctx, ggml_tensor* x) {
         ggml_tensor* w = nullptr;
         ggml_tensor* b = nullptr;
 
@@ -3865,7 +3865,7 @@ public:
         : hidden_size(hidden_size),
           eps(eps) {}
 
-    ggml_tensor* forward(GGMLRunnerContext* ctx, ggml_tensor* x) override {
+    ggml_tensor* forward(GGMLRunnerContext* ctx, ggml_tensor* x) {
         ggml_tensor* w = params["weight"];
         if (ctx->weight_adapter) {
             w = ctx->weight_adapter->patch_weight(ctx->ggml_ctx, ctx->backend, w, prefix + "weight");

src/model/common/block.hpp

@@ -450,7 +450,7 @@ protected:
     int64_t context_dim = 768;  // hidden_size, 1024 for VERSION_SD2
     bool use_linear     = false;
 
-    void init_params(ggml_context* ctx, const String2TensorStorage& tensor_storage_map = {}, const std::string prefix = "") override {
+    void init_params(ggml_context* ctx, const String2TensorStorage& tensor_storage_map = {}, const std::string prefix = "") {
         auto iter = tensor_storage_map.find(prefix + "proj_out.weight");
         if (iter != tensor_storage_map.end()) {
             int64_t inner_dim = n_head * d_head;

src/model/te/llm.hpp

@@ -285,7 +285,7 @@ namespace LLM {
                    bool add_unit_offset = false)
             : hidden_size(hidden_size), eps(eps), add_unit_offset(add_unit_offset) {}
 
-        ggml_tensor* forward(GGMLRunnerContext* ctx, ggml_tensor* x) override {
+        ggml_tensor* forward(GGMLRunnerContext* ctx, ggml_tensor* x) {
             ggml_tensor* w = params["weight"];
             if (ctx->weight_adapter) {
                 w = ctx->weight_adapter->patch_weight(ctx->ggml_ctx, ctx->backend, w, prefix + "weight");

src/model_io/gguf_io.h

@@ -21,7 +21,7 @@ bool write_gguf_file(const std::string& file_path,
 class GGUFStreamingWriter : public StreamingModelWriter {
 public:
     GGUFStreamingWriter() = default;
-    ~GGUFStreamingWriter() override;
+    ~GGUFStreamingWriter();
 
     bool write_metadata(const std::string& file_path,
                         const std::vector<TensorWritePlan>& tensors,

src/runtime/denoiser.hpp

@@ -1153,9 +1153,9 @@ struct CompVisDenoiser : public Denoiser {
         return {c_skip, c_out, c_in};
     }
 
-    sd::Tensor<float> noise_scaling(float sigma,
-                                    const sd::Tensor<float>& noise,
-                                    const sd::Tensor<float>& latent) override {
+    virtual sd::Tensor<float> noise_scaling(float sigma,
+                                            const sd::Tensor<float>& noise,
+                                            const sd::Tensor<float>& latent) override {
         GGML_ASSERT(noise.numel() == latent.numel());
         return latent + noise * sigma;
     }
@@ -1165,7 +1165,7 @@ struct CompVisDenoiser : public Denoiser {
         return latent;
     }
 
-    float noise_level_to_sigma(float noise_level) override {
+    float noise_level_to_sigma(float noise_level) {
         return noise_level / (1.0f - noise_level);
     }
 };
@@ -1256,7 +1256,7 @@ struct DiscreteFlowDenoiser : public Denoiser {
         return latent * (1.0f / (1.0f - sigma));
     }
 
-    float noise_level_to_sigma(float noise_level) override {
+    float noise_level_to_sigma(float noise_level) {
         return noise_level;
     }
 };
@@ -1396,7 +1396,7 @@ struct MiniT2IFlowDenoiser : public Denoiser {
         return latent;
     }
 
-    float noise_level_to_sigma(float noise_level) override {
+    float noise_level_to_sigma(float noise_level) {
         SD_UNUSED(noise_level);
         return 1.0f;
     }