Self-evolving Context Database for AI Agents. Unify Agent Memory, Knowledge RAG and Skills.
Python agent-contextagent-memoryagent-pluginsagentic-ragcontext-databasedsh-pluginknowledge-basememorymemory-managementself-evolvingvector-database
fix(volcengine): forward model request headers (#2909)
Qin Haojie07708225cbac
diff --git a/bot/vikingbot/providers/vlm_adapter.py b/bot/vikingbot/providers/vlm_adapter.pyindex b3e8b585c..07ccdb69f 100644--- a/bot/vikingbot/providers/vlm_adapter.py+++ b/bot/vikingbot/providers/vlm_adapter.py@@ -150,6 +150,9 @@ class VLMProviderAdapter(LLMProvider):"stream": True,"stream_options": {"include_usage": True},}+ extra_headers = getattr(self._vlm, "extra_headers", None)+ if extra_headers:+ kwargs["extra_headers"] = extra_headersif tools:kwargs["tools"] = toolskwargs["tool_choice"] = "auto"diff --git a/openviking/models/embedder/volcengine_embedders.py b/openviking/models/embedder/volcengine_embedders.pyindex df4c5b18b..78530dc56 100644--- a/openviking/models/embedder/volcengine_embedders.py+++ b/openviking/models/embedder/volcengine_embedders.py@@ -20,6 +20,16 @@ from openviking.telemetry import get_current_telemetryfrom openviking.utils.async_client_cache import LoopScopedAsyncClientCachefrom openviking_cli.utils.logger import default_logger as logger+VOLCENGINE_CLIENT_REQUEST_ID_HEADER = "X-Client-Request-Id"+VOLCENGINE_CLIENT_REQUEST_ID = "ToB-direct,OpenViking_Service,openviking-service_cn-beijing"+++def _build_volcengine_headers(extra_headers: Optional[Dict[str, str]]) -> Dict[str, str]:+ headers = dict(extra_headers or {})+ if not any(k.lower() == VOLCENGINE_CLIENT_REQUEST_ID_HEADER.lower() for k in headers):+ headers[VOLCENGINE_CLIENT_REQUEST_ID_HEADER] = VOLCENGINE_CLIENT_REQUEST_ID+ return headers+def to_multimodal_input(content: "EmbeddingInput") -> List[Dict[str, Any]]:"""Normalize an embedding input into the content-parts payload the Volcengine@@ -81,6 +91,7 @@ class VolcengineDenseEmbedder(DenseEmbedderBase):api_base: Optional[str] = None,dimension: Optional[int] = None,input_type: str = "multimodal",+ extra_headers: Optional[Dict[str, str]] = None,config: Optional[Dict[str, Any]] = None,):"""Initialize Volcengine Dense Embedder@@ -103,6 +114,7 @@ class VolcengineDenseEmbedder(DenseEmbedderBase):self.api_base = api_base or "https://ark.cn-beijing.volces.com/api/v3"self.dimension = dimensionself.input_type = input_type+ self.extra_headers = _build_volcengine_headers(extra_headers)if not self.api_key:raise ValueError("api_key is required")@@ -181,14 +193,20 @@ class VolcengineDenseEmbedder(DenseEmbedderBase):if self.input_type == "multimodal":# Use multimodal embeddings APIresponse = self.client.multimodal_embeddings.create(- input=to_multimodal_input(content), model=self.model_name+ input=to_multimodal_input(content),+ model=self.model_name,+ extra_headers=self.extra_headers,)self._update_telemetry_token_usage(response)vector = response.data.embeddingelse:# Use text embeddings API (text-only)text = extract_text_from_content(content)- response = self.client.embeddings.create(input=text, model=self.model_name)+ response = self.client.embeddings.create(+ input=text,+ model=self.model_name,+ extra_headers=self.extra_headers,+ )self._update_telemetry_token_usage(response)vector = response.data[0].embedding@@ -215,13 +233,19 @@ class VolcengineDenseEmbedder(DenseEmbedderBase):async def _embed_call() -> EmbedResult:if self.input_type == "multimodal":response = await client.multimodal_embeddings.create(- input=to_multimodal_input(content), model=self.model_name+ input=to_multimodal_input(content),+ model=self.model_name,+ extra_headers=self.extra_headers,)self._update_telemetry_token_usage(response)vector = response.data.embeddingelse:text = extract_text_from_content(content)- response = await client.embeddings.create(input=text, model=self.model_name)+ response = await client.embeddings.create(+ input=text,+ model=self.model_name,+ extra_headers=self.extra_headers,+ )self._update_telemetry_token_usage(response)vector = response.data[0].embedding@@ -284,6 +308,7 @@ class VolcengineSparseEmbedder(SparseEmbedderBase):api_key: Optional[str] = None,api_base: Optional[str] = None,input_type: str = "multimodal",+ extra_headers: Optional[Dict[str, str]] = None,config: Optional[Dict[str, Any]] = None,):"""Initialize Volcengine Sparse Embedder@@ -304,6 +329,7 @@ class VolcengineSparseEmbedder(SparseEmbedderBase):self.api_key = api_keyself.api_base = api_baseself.input_type = input_type+ self.extra_headers = _build_volcengine_headers(extra_headers)if not self.api_key:raise ValueError("api_key is required")@@ -365,6 +391,7 @@ class VolcengineSparseEmbedder(SparseEmbedderBase):input=to_multimodal_input(content),model=self.model_name,sparse_embedding={"type": "enabled"},+ extra_headers=self.extra_headers,)self._update_telemetry_token_usage(response)item = response.data@@ -393,6 +420,7 @@ class VolcengineSparseEmbedder(SparseEmbedderBase):input=to_multimodal_input(content),model=self.model_name,sparse_embedding={"type": "enabled"},+ extra_headers=self.extra_headers,)self._update_telemetry_token_usage(response)item = response.data@@ -461,6 +489,7 @@ class VolcengineHybridEmbedder(HybridEmbedderBase):api_base: Optional[str] = None,dimension: Optional[int] = None,input_type: str = "multimodal",+ extra_headers: Optional[Dict[str, str]] = None,config: Optional[Dict[str, Any]] = None,):"""Initialize Volcengine Hybrid Embedder@@ -482,6 +511,7 @@ class VolcengineHybridEmbedder(HybridEmbedderBase):self.api_base = api_baseself.dimension = dimensionself.input_type = input_type+ self.extra_headers = _build_volcengine_headers(extra_headers)if not self.api_key:raise ValueError("api_key is required")@@ -550,6 +580,7 @@ class VolcengineHybridEmbedder(HybridEmbedderBase):input=to_multimodal_input(content),model=self.model_name,sparse_embedding={"type": "enabled"},+ extra_headers=self.extra_headers,)self._update_telemetry_token_usage(response)item = response.data@@ -582,6 +613,7 @@ class VolcengineHybridEmbedder(HybridEmbedderBase):input=to_multimodal_input(content),model=self.model_name,sparse_embedding={"type": "enabled"},+ extra_headers=self.extra_headers,)self._update_telemetry_token_usage(response)item = response.datadiff --git a/openviking/models/vlm/backends/volcengine_vlm.py b/openviking/models/vlm/backends/volcengine_vlm.pyindex 4066e97a5..a5eda60b6 100644--- a/openviking/models/vlm/backends/volcengine_vlm.py+++ b/openviking/models/vlm/backends/volcengine_vlm.py@@ -17,6 +17,16 @@ from .openai_vlm import OpenAIVLMlogger = get_logger(__name__)+VOLCENGINE_CLIENT_REQUEST_ID_HEADER = "X-Client-Request-Id"+VOLCENGINE_CLIENT_REQUEST_ID = "ToB-direct,OpenViking_Service,openviking-service_cn-beijing"+++def _build_volcengine_headers(extra_headers: Optional[Dict[str, str]]) -> Dict[str, str]:+ headers = dict(extra_headers or {})+ if not any(k.lower() == VOLCENGINE_CLIENT_REQUEST_ID_HEADER.lower() for k in headers):+ headers[VOLCENGINE_CLIENT_REQUEST_ID_HEADER] = VOLCENGINE_CLIENT_REQUEST_ID+ return headers+class VolcEngineVLM(OpenAIVLM):"""VolcEngine VLM backend with Chat Completions API support."""@@ -25,6 +35,7 @@ class VolcEngineVLM(OpenAIVLM):super().__init__(config)self._sync_client = Noneself.provider = "volcengine"+ self.extra_headers = _build_volcengine_headers(self.extra_headers)if not self.api_base:self.api_base = "https://ark.cn-beijing.volces.com/api/v3"@@ -82,6 +93,7 @@ class VolcEngineVLM(OpenAIVLM):api_key=self.api_key,base_url=self.api_base,timeout=self.timeout,+ max_retries=0,)return self._sync_client@@ -97,6 +109,7 @@ class VolcEngineVLM(OpenAIVLM):api_key=self.api_key,base_url=self.api_base,timeout=self.timeout,+ max_retries=0,)def get_completion(@@ -115,6 +128,7 @@ class VolcEngineVLM(OpenAIVLM):"messages": kwargs_messages,"temperature": self.temperature,"thinking": {"type": "disabled" if not effective_thinking else "enabled"},+ "extra_headers": self.extra_headers,}max_tokens = self.max_tokens or 32768kwargs["max_tokens"] = max_tokens@@ -149,6 +163,7 @@ class VolcEngineVLM(OpenAIVLM):"messages": kwargs_messages,"temperature": self.temperature,"thinking": {"type": "disabled" if not effective_thinking else "enabled"},+ "extra_headers": self.extra_headers,}max_tokens = self.max_tokens or 32768kwargs["max_tokens"] = max_tokens@@ -329,6 +344,7 @@ class VolcEngineVLM(OpenAIVLM):"messages": kwargs_messages,"temperature": self.temperature,"thinking": {"type": "disabled" if not effective_thinking else "enabled"},+ "extra_headers": self.extra_headers,}max_tokens = self.max_tokens or 32768kwargs["max_tokens"] = max_tokens@@ -372,6 +388,7 @@ class VolcEngineVLM(OpenAIVLM):"messages": kwargs_messages,"temperature": self.temperature,"thinking": {"type": "disabled" if not effective_thinking else "enabled"},+ "extra_headers": self.extra_headers,}max_tokens = self.max_tokens or 32768kwargs["max_tokens"] = max_tokensdiff --git a/openviking_cli/utils/config/embedding_config.py b/openviking_cli/utils/config/embedding_config.pyindex 448ab0d00..e6687c487 100644--- a/openviking_cli/utils/config/embedding_config.py+++ b/openviking_cli/utils/config/embedding_config.py@@ -783,6 +783,7 @@ class EmbeddingConfig(BaseModel):"dimension": cfg.dimension,"input_type": cfg.input,"config": dict(runtime_config),+ **({"extra_headers": cfg.extra_headers} if cfg.extra_headers else {}),},),("volcengine", "sparse"): (@@ -792,6 +793,7 @@ class EmbeddingConfig(BaseModel):"api_key": cfg.api_key,"api_base": cfg.api_base,"config": dict(runtime_config),+ **({"extra_headers": cfg.extra_headers} if cfg.extra_headers else {}),},),("volcengine", "hybrid"): (@@ -803,6 +805,7 @@ class EmbeddingConfig(BaseModel):"dimension": cfg.dimension,"input_type": cfg.input,"config": dict(runtime_config),+ **({"extra_headers": cfg.extra_headers} if cfg.extra_headers else {}),},),("vikingdb", "dense"): (