feat: support 16 channel tae (taesd/taef1) (#527)

Author: stduhpf

stable-diffusion.cpp

@@ -360,7 +360,7 @@ class StableDiffusionGGML {
                 first_stage_model->alloc_params_buffer();
                 first_stage_model->get_param_tensors(tensors, "first_stage_model");
             } else {
-                tae_first_stage = std::make_shared<TinyAutoEncoder>(backend, model_loader.tensor_storages_types, "decoder.layers", vae_decode_only);
+                tae_first_stage = std::make_shared<TinyAutoEncoder>(backend, model_loader.tensor_storages_types, "decoder.layers", vae_decode_only, version);
             }
             // first_stage_model->get_param_tensors(tensors, "first_stage_model.");
 

tae.hpp

@@ -62,7 +62,8 @@ class TinyEncoder : public UnaryBlock {
     int num_blocks  = 3;
 
 public:
-    TinyEncoder() {
+    TinyEncoder(int z_channels = 4)
+        : z_channels(z_channels) {
         int index                       = 0;
         blocks[std::to_string(index++)] = std::shared_ptr<GGMLBlock>(new Conv2d(in_channels, channels, {3, 3}, {1, 1}, {1, 1}));
         blocks[std::to_string(index++)] = std::shared_ptr<GGMLBlock>(new TAEBlock(channels, channels));
@@ -106,7 +107,10 @@ class TinyDecoder : public UnaryBlock {
     int num_blocks   = 3;
 
 public:
-    TinyDecoder(int index = 0) {
+    TinyDecoder(int z_channels = 4)
+        : z_channels(z_channels) {
+        int index = 0;
+
         blocks[std::to_string(index++)] = std::shared_ptr<GGMLBlock>(new Conv2d(z_channels, channels, {3, 3}, {1, 1}, {1, 1}));
         index++;  // nn.ReLU()
 
@@ -163,12 +167,16 @@ class TAESD : public GGMLBlock {
     bool decode_only;
 
 public:
-    TAESD(bool decode_only = true)
+    TAESD(bool decode_only = true, SDVersion version = VERSION_SD1)
         : decode_only(decode_only) {
-        blocks["decoder.layers"] = std::shared_ptr<GGMLBlock>(new TinyDecoder());
+        int z_channels = 4;
+        if (sd_version_is_dit(version)) {
+            z_channels = 16;
+        }
+        blocks["decoder.layers"] = std::shared_ptr<GGMLBlock>(new TinyDecoder(z_channels));
 
         if (!decode_only) {
-            blocks["encoder.layers"] = std::shared_ptr<GGMLBlock>(new TinyEncoder());
+            blocks["encoder.layers"] = std::shared_ptr<GGMLBlock>(new TinyEncoder(z_channels));
         }
     }
 
@@ -190,9 +198,10 @@ struct TinyAutoEncoder : public GGMLRunner {
     TinyAutoEncoder(ggml_backend_t backend,
                     std::map<std::string, enum ggml_type>& tensor_types,
                     const std::string prefix,
-                    bool decoder_only = true)
+                    bool decoder_only = true,
+                    SDVersion version = VERSION_SD1)
         : decode_only(decoder_only),
-          taesd(decode_only),
+          taesd(decode_only, version),
           GGMLRunner(backend) {
         taesd.init(params_ctx, tensor_types, prefix);
     }