From 968f3e379c292d395e600dfe5dfbb1e21f905d90 Mon Sep 17 00:00:00 2001
From: xutianle <24210290017@m.fudan.edu.cn>
Date: Wed, 9 Sep 2026 11:33:38 +0800
Subject: [PATCH] [Bugfix][CosyVoice3] Resolve hash snapshot pipeline (#6896)

Signed-off-by: xutianle <xutianle@fudan.edu.cn>
---
 tests/config/test_config_factory.py | 38 +++++++++++++++++++++++++++++
 vllm_omni/deploy/cosyvoice3.yaml    |  1 +
 2 files changed, 39 insertions(+)

diff --git a/tests/config/test_config_factory.py b/tests/config/test_config_factory.py
index 6cb328bc5..51c7d549e 100644
--- a/tests/config/test_config_factory.py
+++ b/tests/config/test_config_factory.py
@@ -932,6 +932,44 @@ class TestPipelineRegistration:
         assert isinstance(omni_config, VllmOmniConfig)
         assert omni_config.stage_by_id(0).diffusion_config.model == "fake/model"
 
+    def test_cosyvoice3_deploy_resolves_hash_snapshot_without_hf_metadata(self, tmp_path):
+        snapshot = tmp_path / "models--FunAudioLLM--Fun-CosyVoice3-0.5B-2512" / "snapshots" / ("a" * 40)
+        snapshot.mkdir(parents=True)
+        (snapshot / "config.json").write_text("{}\n", encoding="utf-8")
+        model = str(snapshot)
+        deploy_path = get_deploy_config_path("cosyvoice3.yaml")
+
+        assert StageConfigFactory.try_infer_model_type(model, trust_remote_code=False) is None
+        assert StageConfigFactory.get_pipeline_config(model, trust_remote_code=False) is None
+
+        omni_config = StageConfigFactory.create_from_model(
+            model,
+            trust_remote_code=False,
+            cli_overrides={},
+            deploy_config_path=deploy_path,
+        )
+        legacy_configs, _ = StageConfigFactory.create_legacy_stage_configs_from_model(
+            model,
+            trust_remote_code=False,
+            cli_overrides={},
+            deploy_config_path=deploy_path,
+        )
+
+        assert omni_config is not None
+        assert omni_config.pipeline_config.model_type == "cosyvoice3"
+        assert [
+            (stage.stage_id, stage.stage_type, stage.worker_type, stage.model_stage)
+            for stage in omni_config.stage_configs
+        ] == [
+            (0, StageType.LLM, "ar", "cosyvoice3_talker"),
+            (1, StageType.LLM, "generation", "cosyvoice3_code2wav"),
+        ]
+        assert all(stage.model_config.model == model for stage in omni_config.stage_configs)
+        assert all(stage.stage_type is not StageType.DIFFUSION for stage in omni_config.stage_configs)
+
+        assert legacy_configs is not None
+        assert [stage.stage_type for stage in legacy_configs] == [StageType.LLM, StageType.LLM]
+
     def test_pipeline_registration(self, clean_pipeline_registry):
         """Ensure that we can register and create a custom pipeline config."""
         new_model_type = "new_model_type"
diff --git a/vllm_omni/deploy/cosyvoice3.yaml b/vllm_omni/deploy/cosyvoice3.yaml
index 551b42af6..09fd83290 100644
--- a/vllm_omni/deploy/cosyvoice3.yaml
+++ b/vllm_omni/deploy/cosyvoice3.yaml
@@ -23,6 +23,7 @@
 #   On a slower GPU class, lower max_num_seqs (e.g. 4) and/or codec_chunk_frames
 #   first. Raising connector_get_sleep_s also relaxes the consumer's polling
 #   cadence, but trades against TTFP — keep it as a last resort.
+pipeline: cosyvoice3
 async_chunk: true
 connectors:
   connector_of_shared_memory: