From 0443587a570334843d56f94e50e4358e40b254a4 Mon Sep 17 00:00:00 2001
From: gaoziyuan <88373061+gzy19990617@users.noreply.github.com>
Date: Mon, 4 Aug 2025 16:34:07 +0800
Subject: [PATCH] =?UTF-8?q?=E3=80=90Feature=E3=80=91support=20qwen3=20name?=
 =?UTF-8?q?=5Fmapping=20(#3179)?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

* add fd plugins && rm model_classed

* fix reviews

* add docs

* fix

* fix unitest ci

* support qwen3 name_mapping
---
 fastdeploy/rl/rollout_model.py | 22 +++++++++++++++++++++-
 1 file changed, 21 insertions(+), 1 deletion(-)

diff --git a/fastdeploy/rl/rollout_model.py b/fastdeploy/rl/rollout_model.py
index 6d0974a5f..11701c0e0 100644
--- a/fastdeploy/rl/rollout_model.py
+++ b/fastdeploy/rl/rollout_model.py
@@ -429,4 +429,24 @@ class Qwen3ForCausalLMRL(Qwen3ForCausalLM, BaseRLModel):
         return "Qwen3ForCausalLMRL"
 
     def get_name_mappings_to_training(self, trainer_degree=None) -> Dict[str, str]:
-        pass
+        # Prepare placeholders
+        place_holders = ["weight"]
+
+        # Initialize mapping dictionary
+        self._update_base_mappings("model")
+        base_name = "model.layers"
+
+        # Helper function to add layer mappings
+        def _add_layer_mappings(layer_idx):
+            # FFN mappings
+            for ph in place_holders:
+                self.infer_to_train_mapping[f"{base_name}.{layer_idx}.mlp.up_gate_proj.{ph}"] = (
+                    f"{base_name}.{layer_idx}.mlp.gate_up_fused_proj.{ph}"
+                )
+
+        for layer_idx in range(self.fd_config.model_config.num_hidden_layers):
+            _add_layer_mappings(layer_idx)
+
+        self._complete_missing_mappings()
+
+        return self.infer_to_train_mapping