diff --git a/examples/common/common.hpp b/examples/common/common.hpp
index bf38379d2..b9ac7edc1 100644
--- a/examples/common/common.hpp
+++ b/examples/common/common.hpp
@@ -863,7 +863,7 @@ static bool is_absolute_path(const std::string& p) {
 
 struct SDGenerationParams {
     std::string prompt;
-    std::string prompt_with_lora; // for metadata record only
+    std::string prompt_with_lora;  // for metadata record only
     std::string negative_prompt;
     int clip_skip   = -1;  // <= 0 represents unspecified
     int width       = 512;
diff --git a/flux.hpp b/flux.hpp
index 1df2874ae..7ce263569 100644
--- a/flux.hpp
+++ b/flux.hpp
@@ -744,6 +744,8 @@ namespace Flux {
         int64_t nerf_mlp_ratio   = 4;
         int64_t nerf_depth       = 4;
         int64_t nerf_max_freqs   = 8;
+        bool use_x0              = false;
+        bool use_patch_size_32   = false;
     };
 
     struct FluxParams {
@@ -781,7 +783,7 @@ namespace Flux {
         Flux(FluxParams params)
             : params(params) {
             if (params.version == VERSION_CHROMA_RADIANCE) {
-                std::pair<int, int> kernel_size = {(int)params.patch_size, (int)params.patch_size};
+                std::pair<int, int> kernel_size = {16, 16};
                 std::pair<int, int> stride      = kernel_size;
 
                 blocks["img_in_patch"] = std::make_shared<Conv2d>(params.in_channels,
@@ -1044,6 +1046,15 @@ namespace Flux {
             return img;
         }
 
+        struct ggml_tensor* _apply_x0_residual(GGMLRunnerContext* ctx,
+                                               struct ggml_tensor* predicted,
+                                               struct ggml_tensor* noisy,
+                                               struct ggml_tensor* timesteps) {
+            auto x = ggml_sub(ctx->ggml_ctx, noisy, predicted);
+            x      = ggml_div(ctx->ggml_ctx, x, timesteps);
+            return x;
+        }
+
         struct ggml_tensor* forward_chroma_radiance(GGMLRunnerContext* ctx,
                                                     struct ggml_tensor* x,
                                                     struct ggml_tensor* timestep,
@@ -1068,6 +1079,13 @@ namespace Flux {
             auto img      = pad_to_patch_size(ctx->ggml_ctx, x);
             auto orig_img = img;
 
+            if (params.chroma_radiance_params.use_patch_size_32) {
+                // It's supposed to be using GGML_SCALE_MODE_NEAREST, but this seems more stable
+                // Maybe the implementation of nearest-neighbor interpolation in ggml behaves differently than the one in PyTorch?
+                // img = F.interpolate(img, size=(H//2, W//2), mode="nearest")
+                img = ggml_interpolate(ctx->ggml_ctx, img, W / 2, H / 2, C, x->ne[3], GGML_SCALE_MODE_BILINEAR);
+            }
+
             auto img_in_patch = std::dynamic_pointer_cast<Conv2d>(blocks["img_in_patch"]);
 
             img = img_in_patch->forward(ctx, img);                                                       // [N, hidden_size, H/patch_size, W/patch_size]
@@ -1104,6 +1122,10 @@ namespace Flux {
 
             out = nerf_final_layer_conv->forward(ctx, img_dct);  // [N, C, H, W]
 
+            if (params.chroma_radiance_params.use_x0) {
+                out = _apply_x0_residual(ctx, out, orig_img, timestep);
+            }
+
             return out;
         }
 
@@ -1290,6 +1312,15 @@ namespace Flux {
                     // not schnell
                     flux_params.guidance_embed = true;
                 }
+                if (tensor_name.find("__x0__") != std::string::npos) {
+                    LOG_DEBUG("using x0 prediction");
+                    flux_params.chroma_radiance_params.use_x0 = true;
+                }
+                if (tensor_name.find("__32x32__") != std::string::npos) {
+                    LOG_DEBUG("using patch size 32 prediction");
+                    flux_params.chroma_radiance_params.use_patch_size_32 = true;
+                    flux_params.patch_size                               = 32;
+                }
                 if (tensor_name.find("distilled_guidance_layer.in_proj.weight") != std::string::npos) {
                     // Chroma
                     flux_params.is_chroma = true;
diff --git a/model.cpp b/model.cpp
index 0480efefb..74bcdcfe2 100644
--- a/model.cpp
+++ b/model.cpp
@@ -1732,6 +1732,13 @@ bool ModelLoader::save_to_gguf_file(const std::string& file_path, ggml_type type
         // tensor_storage.ne[0], tensor_storage.ne[1], tensor_storage.ne[2], tensor_storage.ne[3],
         // tensor->n_dims, tensor->ne[0], tensor->ne[1], tensor->ne[2], tensor->ne[3]);
 
+        if (!tensor->data) {
+            GGML_ASSERT(ggml_nelements(tensor) == 0);
+            // avoid crashing the gguf writer by setting a dummy pointer for zero-sized tensors
+            LOG_DEBUG("setting dummy pointer for zero-sized tensor %s", name.c_str());
+            tensor->data = ggml_get_mem_buffer(ggml_ctx);
+        }
+
         *dst_tensor = tensor;
 
         gguf_add_tensor(gguf_ctx, tensor);
diff --git a/stable-diffusion.cpp b/stable-diffusion.cpp
index 2cb588213..ad8fd5c74 100644
--- a/stable-diffusion.cpp
+++ b/stable-diffusion.cpp
@@ -695,6 +695,8 @@ class StableDiffusionGGML {
         if (stacked_id) {
             ignore_tensors.insert("pmid.unet.");
         }
+        ignore_tensors.insert("model.diffusion_model.__x0__");
+        ignore_tensors.insert("model.diffusion_model.__32x32__");
 
         if (vae_decode_only) {
             ignore_tensors.insert("first_stage_model.encoder");
@@ -829,6 +831,7 @@ class StableDiffusionGGML {
                     }
                 } else if (sd_version_is_flux(version)) {
                     pred_type = FLUX_FLOW_PRED;
+
                     if (flow_shift == INFINITY) {
                         flow_shift = 1.0f;  // TODO: validate
                         for (const auto& [name, tensor_storage] : tensor_storage_map) {