[Model] Qwen3.5 dense and MoE support (no vision) (#19435)

* Unified delta net handling

* Remove old methods.

* Refactor and optimize

* Adapt autoregressive version from @ymcki

* Change to decay mask approach

* Fix bad permute

* Qwen 3.5 support

* Apply suggestions from code review

Co-authored-by: Sigbjørn Skjæret <sigbjorn.skjaeret@scala.com>

* Further fixes

* Use inheritance, remove unneeded conts

* Not like this!

* Remove ggml.h explicit import

* Remove transformers, fix the views

* ACTUALLY fix views, make super calls explicit in conversion.

* Fix conversion again

* Remove extra ggml.h imports

---------

Co-authored-by: Sigbjørn Skjæret <sigbjorn.skjaeret@scala.com>

gguf-py/gguf/tensor_mapping.py

@@ -228,6 +228,7 @@ class TensorNameMap:
             "transformer_encoder.{bid}.qkv",                                       # neobert
             "layers.{bid}.attn.Wqkv",                                              # modern-bert
             "model.layers.{bid}.self_attn.language_expert_query_key_value",        # cogvlm
+            "model.layers.{bid}.linear_attn.in_proj_qkv",                          # qwen3.5
         ),
 
         # Attention query
@@ -358,8 +359,9 @@ class TensorNameMap:
         ),
 
         MODEL_TENSOR.ATTN_GATE: (
-            "model.layers.{bid}.self_attn.gate_proj", # afmoe
-            "model.layers.{bid}.self_attn.g_proj",    # step3.5 head-wise attention gate
+            "model.layers.{bid}.self_attn.gate_proj",   # afmoe
+            "model.layers.{bid}.self_attn.g_proj",      # step3.5 head-wise attention gate
+            "model.layers.{bid}.linear_attn.in_proj_z", # qwen3.5
         ),
 
         # Feed-forward norm