refactor: centralize CLIP prefix conversion (#1837)

examples/common/common.cpp

@@ -1008,7 +1008,7 @@ ArgOptions SDGenerationParams::get_options() {
          &hires_upscaler},
         {"",
          "--extra-sample-args",
-         "extra sampler/scheduler/guidance args, key=value list. CFG supports guidance_schedule; APG supports apg_eta, apg_momentum, apg_norm_threshold, apg_norm_threshold_smoothing; SLG supports slg_uncond; lcm supports noise_clip_std, noise_scale_start, noise_scale_end; flux supports base_shift, max_shift; ltx2 supports max_shift, base_shift, stretch, terminal; euler_ge supports gamma;; logit_normal supports mu, std, logsnr_min, logsnr_max, resolution_aware",
+         "extra sampler/scheduler/guidance args, key=value list. CFG supports guidance_schedule; APG supports apg_eta, apg_momentum, apg_norm_threshold, apg_norm_threshold_smoothing; SLG supports slg_uncond; lcm supports noise_clip_std, noise_scale_start, noise_scale_end; flux supports base_shift, max_shift; ltx2 supports max_shift, base_shift, stretch, terminal; euler_ge supports gamma; beta scheduler supports alpha, beta; logit_normal supports mu, std, logsnr_min, logsnr_max, resolution_aware",
          (int)',',
          &extra_sample_args},
         {"",

src/extensions/photomaker_extension.cpp

@@ -56,7 +56,7 @@ tokenize_photomaker_trigger(FrozenCLIPEmbedderWithCustomWords& clip_conditioner,
                                           true);
     std::vector<bool> class_token_mask;
     for (int i = 0; i < tokens.size(); i++) {
-        class_token_mask.push_back(class_idx + 1 <= i && i < class_idx + 1 + trigger_token_count);
+        class_token_mask.push_back(class_idx >= 0 && class_idx + 1 <= i && i < class_idx + 1 + trigger_token_count);
     }
 
     return std::make_tuple(tokens, weights, class_token_mask);

src/name_conversion.cpp

@@ -1449,12 +1449,21 @@ std::string convert_tensor_name(std::string name, SDVersion version) {
         {"te2.", "cond_stage_model.1.transformer."},
         {"te1.", "cond_stage_model.transformer."},
         {"te3.", "text_encoders.t5xxl.transformer."},
+        {"clip_vision.", "cond_stage_model.transformer."},
     };
 
     if (sd_version_is_flux(version)) {
         prefix_map["te1."] = "text_encoders.clip_l.transformer.";
     }
 
+    if (sd_version_is_unet(version)) {
+        prefix_map["clip_l."] = "cond_stage_model.transformer.";
+        prefix_map["clip_g."] = "cond_stage_model.1.transformer.";
+    } else {
+        prefix_map["clip_l."] = "text_encoders.clip_l.transformer.";
+        prefix_map["clip_g."] = "text_encoders.clip_g.transformer.";
+    }
+
     replace_with_prefix_map(name, prefix_map);
 
     if (sd_version_is_boogu_image(version) || sd_version_is_krea2(version) || sd_version_is_mage_flow(version)) {

src/runtime/denoiser.hpp

@@ -306,8 +306,33 @@ struct KarrasScheduler : SigmaScheduler {
 };
 
 struct BetaScheduler : SigmaScheduler {
-    static constexpr double alpha = 0.6;
+    double alpha = 0.6;
-    static constexpr double beta  = 0.6;
+    double beta  = 0.6;
+
+    explicit BetaScheduler(const char* extra_sample_args = nullptr) {
+        parse_extra_sample_args(extra_sample_args);
+        LOG_DEBUG("Beta scheduler: alpha=%.4f, beta=%.4f", alpha, beta);
+    }
+
+    void parse_extra_sample_args(const char* extra_sample_args) {
+        for (const auto& [key, value] : parse_key_value_args(extra_sample_args, "beta scheduler arg")) {
+            if (key == "alpha") {
+                float parsed;
+                if (!parse_strict_float(value, parsed) || parsed <= 0.0) {
+                    LOG_WARN("ignoring invalid beta scheduler arg '%s=%s'", key.c_str(), value.c_str());
+                } else {
+                    alpha = static_cast<double>(parsed);
+                }
+            } else if (key == "beta") {
+                float parsed;
+                if (!parse_strict_float(value, parsed) || parsed <= 0.0) {
+                    LOG_WARN("ignoring invalid beta scheduler arg '%s=%s'", key.c_str(), value.c_str());
+                } else {
+                    beta = static_cast<double>(parsed);
+                }
+            }
+        }
+    }
 
     static double log_beta(double a, double b) {
         return std::lgamma(a) + std::lgamma(b) - std::lgamma(a + b);
@@ -1032,7 +1057,7 @@ struct Denoiser {
                 break;
             case BETA_SCHEDULER:
                 LOG_INFO("get_sigmas with Beta scheduler");
-                scheduler = std::make_shared<BetaScheduler>();
+                scheduler = std::make_shared<BetaScheduler>(extra_sample_args);
                 break;
             case EXPONENTIAL_SCHEDULER:
                 LOG_INFO("get_sigmas exponential scheduler");

src/stable-diffusion.cpp

@@ -762,28 +762,23 @@ public:
             }
         }
 
-        bool is_unet = sd_version_is_unet(model_loader.get_sd_version());
-
         if (strlen(SAFE_STR(sd_ctx_params->clip_l_path)) > 0) {
             LOG_INFO("loading clip_l from '%s'", sd_ctx_params->clip_l_path);
-            std::string prefix = is_unet ? "cond_stage_model.transformer." : "text_encoders.clip_l.transformer.";
+            if (!model_loader.init_from_file(sd_ctx_params->clip_l_path, "clip_l.")) {
-            if (!model_loader.init_from_file(sd_ctx_params->clip_l_path, prefix)) {
                 LOG_WARN("loading clip_l from '%s' failed", sd_ctx_params->clip_l_path);
             }
         }
 
         if (strlen(SAFE_STR(sd_ctx_params->clip_g_path)) > 0) {
             LOG_INFO("loading clip_g from '%s'", sd_ctx_params->clip_g_path);
-            std::string prefix = is_unet ? "cond_stage_model.1.transformer." : "text_encoders.clip_g.transformer.";
+            if (!model_loader.init_from_file(sd_ctx_params->clip_g_path, "clip_g.")) {
-            if (!model_loader.init_from_file(sd_ctx_params->clip_g_path, prefix)) {
                 LOG_WARN("loading clip_g from '%s' failed", sd_ctx_params->clip_g_path);
             }
         }
 
         if (strlen(SAFE_STR(sd_ctx_params->clip_vision_path)) > 0) {
             LOG_INFO("loading clip_vision from '%s'", sd_ctx_params->clip_vision_path);
-            std::string prefix = "cond_stage_model.transformer.";
+            if (!model_loader.init_from_file(sd_ctx_params->clip_vision_path, "clip_vision.")) {
-            if (!model_loader.init_from_file(sd_ctx_params->clip_vision_path, prefix)) {
                 LOG_WARN("loading clip_vision from '%s' failed", sd_ctx_params->clip_vision_path);
             }
         }