return filter to save memory (#24125)

Co-authored-by: lvyichen <lvyichen@stepfun.com>

conversion/gemma.py

@@ -798,7 +798,8 @@ class Gemma4VisionAudioModel(MmprojModel):
         # remap audio hparams
         if self.hparams_audio:
             self.hparams_audio["feat_in"] = self.hparams_audio.get("input_feat_size", 128)
-            self.hparams_audio["intermediate_size"] = self.hparams_audio["hidden_size"] * 4
+            if "hidden_size" in self.hparams_audio:
+                self.hparams_audio["intermediate_size"] = self.hparams_audio["hidden_size"] * 4
         else:
             self.has_audio_encoder = False
 
@@ -872,7 +873,7 @@ class Gemma4UnifiedVisionAudioModel(Gemma4VisionAudioModel):
         assert self.hparams_audio is not None
         text_embd_dim = self.hparams_vision["mm_embed_dim"]
         self.hparams_vision["hidden_size"] = text_embd_dim
-        self.hparams_audio["hidden_size"] = text_embd_dim
+        self.hparams_audio["hidden_size"] = self.hparams_audio["audio_embed_dim"]
         # this is a transformer-less vision tower, the params below are redundant but set to avoid error
         self.hparams_vision["intermediate_size"] = 0
         self.hparams_vision["num_layers"] = 0
@@ -897,7 +898,10 @@ class Gemma4UnifiedVisionAudioModel(Gemma4VisionAudioModel):
             # ggml im2col outputs in RR..GG..BB.. (CHW) order, but weight expects RGBRGB.. (HWC).
             # Permute columns so column i aligns with CHW input position i.
             assert self.hparams_vision is not None
-            p = self.hparams_vision["model_patch_size"]
+            if "model_patch_size" in self.hparams_vision:
+                p = self.hparams_vision["model_patch_size"]
+            else:
+                p = self.hparams_vision["patch_size"] * self.hparams_vision["pooling_kernel_size"]
             i = torch.arange(p * p * 3)
             ch  = i // (p * p)
             row = (i % (p * p)) // p
@@ -908,7 +912,10 @@ class Gemma4UnifiedVisionAudioModel(Gemma4VisionAudioModel):
         elif "patch_ln1.weight" in name or "patch_ln1.bias" in name:
             # same permutation for patch_ln1 as patch_dense to align with CHW input order
             assert self.hparams_vision is not None
-            p = self.hparams_vision["model_patch_size"]
+            if "model_patch_size" in self.hparams_vision:
+                p = self.hparams_vision["model_patch_size"]
+            else:
+                p = self.hparams_vision["patch_size"] * self.hparams_vision["pooling_kernel_size"]
             i = torch.arange(p * p * 3)
             ch  = i // (p * p)
             row = (i % (p * p)) // p

src/llama-model.cpp

@@ -2112,6 +2112,15 @@ llama_memory_i * llama_model::create_memory(const llama_memory_params & params,
                         filter = [n_main](int32_t il) { return (uint32_t)il >= n_main; };
                     }
 
+                    if (arch == LLM_ARCH_STEP35 && hparams.nextn_predict_layers > 0) {
+                        const uint32_t n_main = hparams.n_layer - hparams.nextn_predict_layers;
+                        if (params.ctx_type == LLAMA_CONTEXT_TYPE_MTP) {
+                            filter = [n_main](int32_t il) { return (uint32_t)il >= n_main; };
+                        } else {
+                            filter = [n_main](int32_t il) { return (uint32_t)il <  n_main; };
+                        }
+                    }
+
                     if (hparams.swa_type != LLAMA_SWA_TYPE_NONE) {
                         GGML_ASSERT(hparams.is_swa_any());