feat: change image dimensions requirement for DiT models (#742)

stduhpf · web-flow · commit 59080d3ce12e · 2025-07-28T21:58:17.000+08:00
diff --git a/examples/cli/main.cpp b/examples/cli/main.cpp
@@ -596,13 +596,13 @@ void parse_args(int argc, const char** argv, SDParams& params) {
         exit(1);
     }
 
-    if (params.width <= 0 || params.width % 64 != 0) {
-        fprintf(stderr, "error: the width must be a multiple of 64\n");
+    if (params.width <= 0) {
+        fprintf(stderr, "error: the width must be greater than 0\n");
         exit(1);
     }
 
-    if (params.height <= 0 || params.height % 64 != 0) {
-        fprintf(stderr, "error: the height must be a multiple of 64\n");
+    if (params.height <= 0) {
+        fprintf(stderr, "error: the height must be greater than 0\n");
         exit(1);
     }
 
diff --git a/stable-diffusion.cpp b/stable-diffusion.cpp
@@ -1875,6 +1875,15 @@ ggml_tensor* generate_init_latent(sd_ctx_t* sd_ctx,
 sd_image_t* generate_image(sd_ctx_t* sd_ctx, const sd_img_gen_params_t* sd_img_gen_params) {
     int width  = sd_img_gen_params->width;
     int height = sd_img_gen_params->height;
+    if (sd_version_is_dit(sd_ctx->sd->version)) {
+        if (width % 16 || height % 16) {
+            LOG_ERROR("Image dimensions must be must be a multiple of 16 on each axis for %s models. (Got %dx%d)", model_version_to_str[sd_ctx->sd->version], width, height);
+            return NULL;
+        }
+    } else if (width % 64 || height % 64) {
+        LOG_ERROR("Image dimensions must be must be a multiple of 64 on each axis for %s models. (Got %dx%d)", model_version_to_str[sd_ctx->sd->version], width, height);
+        return NULL;
+    }
     LOG_DEBUG("generate_image %dx%d", width, height);
     if (sd_ctx == NULL || sd_img_gen_params == NULL) {
         return NULL;

Original file line number	Diff line number	Diff line change
`@@ -596,13 +596,13 @@ void parse_args(int argc, const char** argv, SDParams& params) {`
`596`	`596`	`exit(1);`
`597`	`597`	`}`
`598`	`598`
`599`		`- if (params.width <= 0 \|\| params.width % 64 != 0) {`
`600`		`- fprintf(stderr, "error: the width must be a multiple of 64\n");`
	`599`	`+ if (params.width <= 0) {`
	`600`	`+ fprintf(stderr, "error: the width must be greater than 0\n");`
`601`	`601`	`exit(1);`
`602`	`602`	`}`
`603`	`603`
`604`		`- if (params.height <= 0 \|\| params.height % 64 != 0) {`
`605`		`- fprintf(stderr, "error: the height must be a multiple of 64\n");`
	`604`	`+ if (params.height <= 0) {`
	`605`	`+ fprintf(stderr, "error: the height must be greater than 0\n");`
`606`	`606`	`exit(1);`
`607`	`607`	`}`
`608`	`608`