From a9ccd7ba42d0d486541dc79538b370be7e17e91f Mon Sep 17 00:00:00 2001 From: "fangyaozheng@bytedance.com" Date: Mon, 1 Sep 2025 12:12:41 +0800 Subject: [PATCH 1/2] chore(config): add model config yaml --- config.yaml.full | 2 ++ config.yaml.simple | 4 +++- docs/docs/agent.md | 1 + tests/test_agent.py | 18 +++++++++++------ veadk/agent.py | 49 +++++++++++++++++++++++++++++++++++++-------- veadk/consts.py | 15 +++++++++++++- 6 files changed, 73 insertions(+), 16 deletions(-) diff --git a/config.yaml.full b/config.yaml.full index aaf601d0..6c495f97 100644 --- a/config.yaml.full +++ b/config.yaml.full @@ -5,6 +5,8 @@ model: name: doubao-seed-1-6-250615 api_base: https://ark.cn-beijing.volces.com/api/v3/ api_key: + extra_headers: '{"x-is-encrypted": "true"}' + extra_body: '{"caching": {"type": "enabled"}, "thinking": {"type": "disabled"}, "expire_at": 1}' # expire after 1 hour # [optional] for llm-as-a-judge a evaluation judge: name: doubao-seed-1-6-250615 diff --git a/config.yaml.simple b/config.yaml.simple index 9476d107..95587a80 100644 --- a/config.yaml.simple +++ b/config.yaml.simple @@ -3,4 +3,6 @@ model: provider: openai name: doubao-seed-1-6-250615 api_base: https://ark.cn-beijing.volces.com/api/v3/ - api_key: \ No newline at end of file + api_key: + extra_headers: '{"x-is-encrypted": "true"}' + extra_body: '{"caching": {"type": "enabled"}, "thinking": {"type": "disabled"}, "expire_at": 1}' # expire after 1 hour \ No newline at end of file diff --git a/docs/docs/agent.md b/docs/docs/agent.md index 056e7405..88f2dd62 100644 --- a/docs/docs/agent.md +++ b/docs/docs/agent.md @@ -15,6 +15,7 @@ Agent 中主要包括如下属性： | model_provider | str | Agent 中内置模型的提供商，默认从环境变量中获取 | | model_api_base | str | Agent 中内置模型的 API Base，默认从环境变量中获取 | | model_api_key | str | Agent 中内置模型的 API Key，默认从环境变量中获取 | +| model_extra_config | dict | Agent 进行模型请求时的额外参数，Key 值为属性名，Value 值为属性值 | | tools | list | Function call 中的工具列表，既可以是本地工具，也可以是 MCP 工具 | | sub_agents | list | 子 Agent 列表，用于多 Agent 之间交互 | | knowledgebase | KnowledgeBase | 知识库，后端支持本地内存（local）和数据库（opensearch、viking、redis、mysql），通常设置为一个能够检索的向量数据库 | diff --git a/tests/test_agent.py b/tests/test_agent.py index 8ad947a7..2ecf4522 100644 --- a/tests/test_agent.py +++ b/tests/test_agent.py @@ -15,7 +15,7 @@ from google.adk.tools import load_memory from veadk import Agent -from veadk.consts import DEFAULT_MODEL_EXTRA_HEADERS +from veadk.consts import DEFAULT_MODEL_EXTRA_CONFIG from veadk.knowledgebase import KnowledgeBase from veadk.memory.long_term_memory import LongTermMemory from veadk.tools import load_knowledgebase_tool @@ -27,14 +27,17 @@ def test_agent(): long_term_memory = LongTermMemory(backend="local") tracer = OpentelemetryTracer() - model_extra_headers = {"test-header": "test-value"} + extra_config = { + "extra_headers": {"thinking": "test"}, + "extra_body": {"content": "test"}, + } agent = Agent( model_name="test_model_name", model_provider="test_model_provider", model_api_key="test_model_api_key", model_api_base="test_model_api_base", - model_extra_headers=model_extra_headers, + model_extra_config=extra_config, tools=[], sub_agents=[], knowledgebase=knowledgebase, @@ -43,10 +46,13 @@ def test_agent(): serve_url="", ) - model_extra_headers |= DEFAULT_MODEL_EXTRA_HEADERS - assert agent.model.model == f"{agent.model_provider}/{agent.model_name}" - assert agent.model_extra_headers == model_extra_headers + + expected_config = DEFAULT_MODEL_EXTRA_CONFIG.copy() + expected_config["extra_headers"] |= extra_config["extra_headers"] + expected_config["extra_body"] |= extra_config["extra_body"] + + assert agent.model_extra_config == expected_config assert agent.knowledgebase == knowledgebase assert agent.knowledgebase.backend == "local" diff --git a/veadk/agent.py b/veadk/agent.py index 99c50ec0..768f2e8c 100644 --- a/veadk/agent.py +++ b/veadk/agent.py @@ -28,10 +28,10 @@ from veadk.config import getenv from veadk.consts import ( - DEFAULT_MODEL_AGENT_PROVIDER, DEFAULT_MODEL_AGENT_API_BASE, DEFAULT_MODEL_AGENT_NAME, - DEFAULT_MODEL_EXTRA_HEADERS, + DEFAULT_MODEL_AGENT_PROVIDER, + DEFAULT_MODEL_EXTRA_CONFIG, ) from veadk.evaluation import EvalSetRecorder from veadk.knowledgebase import KnowledgeBase @@ -47,6 +47,29 @@ logger = get_logger(__name__) +def _parse_user_model_extra_config() -> dict: + import json + + extra_headers = getenv("MODEL_AGENT_EXTRA_HEADERS", {}, allow_false_values=True) + extra_body = getenv("MODEL_AGENT_EXTRA_BODY", {}, allow_false_values=True) + + if extra_headers: + extra_headers = json.loads(extra_headers) + if extra_body: + extra_body = json.loads(extra_body) + if extra_body.get("expire_at", 0): + import time + + extra_body["expire_at"] = ( + int(time.time()) + int(extra_body["expire_at"]) * 3600 + ) + + return { + "extra_headers": extra_headers, + "extra_body": extra_body, + } + + class Agent(LlmAgent): """LLM-based Agent with Volcengine capabilities.""" @@ -74,7 +97,7 @@ class Agent(LlmAgent): model_api_key: str = Field(default_factory=lambda: getenv("MODEL_AGENT_API_KEY")) """The api key of the model for agent running.""" - model_extra_config: dict = Field(default_factory=dict) + model_extra_config: dict = Field(default_factory=_parse_user_model_extra_config) """The extra config to include in the model requests.""" tools: list[ToolUnion] = [] @@ -101,11 +124,21 @@ class Agent(LlmAgent): def model_post_init(self, __context: Any) -> None: super().model_post_init(None) # for sub_agents init - # add model request source (veadk) in extra headers - if self.model_extra_config and "extra_headers" in self.model_extra_config: - self.model_extra_config["extra_headers"] |= DEFAULT_MODEL_EXTRA_HEADERS - else: - self.model_extra_config["extra_headers"] = DEFAULT_MODEL_EXTRA_HEADERS + # combine user model config with VeADK defaults + headers = DEFAULT_MODEL_EXTRA_CONFIG["extra_headers"].copy() + body = DEFAULT_MODEL_EXTRA_CONFIG["extra_body"].copy() + + if self.model_extra_config: + user_headers = self.model_extra_config.get("extra_headers", {}) + user_body = self.model_extra_config.get("extra_body", {}) + + headers |= user_headers + body |= user_body + + self.model_extra_config |= { + "extra_headers": headers, + "extra_body": body, + } if not self.model: self.model = LiteLlm( diff --git a/veadk/consts.py b/veadk/consts.py index 5c5ac71a..fad5aecc 100644 --- a/veadk/consts.py +++ b/veadk/consts.py @@ -12,12 +12,25 @@ # See the License for the specific language governing permissions and # limitations under the License. +import time + from veadk.version import VERSION DEFAULT_MODEL_AGENT_NAME = "doubao-seed-1-6-250615" DEFAULT_MODEL_AGENT_PROVIDER = "openai" DEFAULT_MODEL_AGENT_API_BASE = "https://ark.cn-beijing.volces.com/api/v3/" -DEFAULT_MODEL_EXTRA_HEADERS = {"veadk-source": "veadk", "veadk-version": VERSION} +DEFAULT_MODEL_EXTRA_CONFIG = { + "extra_headers": { + "x-is-encrypted": "true", + "veadk-source": "veadk", + "veadk-version": VERSION, + }, + "extra_body": { + "caching": {"type": "enabled"}, + "thinking": {"type": "disabled"}, + "expire_at": int(time.time()) + 3600, # expire after 1 hour + }, +} DEFAULT_APMPLUS_OTEL_EXPORTER_ENDPOINT = "http://apmplus-cn-beijing.volces.com:4317" DEFAULT_APMPLUS_OTEL_EXPORTER_SERVICE_NAME = "veadk_tracing" From 0dc53a006802c630ff4f29f26f969c3d3013f00a Mon Sep 17 00:00:00 2001 From: "fangyaozheng@bytedance.com" Date: Mon, 1 Sep 2025 17:45:54 +0800 Subject: [PATCH 2/2] fix comfig.yaml --- config.yaml.full | 5 +++-- config.yaml.simple | 5 +++-- veadk/agent.py | 27 +++------------------------ veadk/consts.py | 8 +++++--- veadk/runner.py | 19 ++++++++++++++++--- 5 files changed, 30 insertions(+), 34 deletions(-) diff --git a/config.yaml.full b/config.yaml.full index 6c495f97..6c0cef57 100644 --- a/config.yaml.full +++ b/config.yaml.full @@ -5,8 +5,9 @@ model: name: doubao-seed-1-6-250615 api_base: https://ark.cn-beijing.volces.com/api/v3/ api_key: - extra_headers: '{"x-is-encrypted": "true"}' - extra_body: '{"caching": {"type": "enabled"}, "thinking": {"type": "disabled"}, "expire_at": 1}' # expire after 1 hour + encrypted: true # true | false + caching: enabled # enabled | disabled + max_llm_calls: 100 # [optional] for llm-as-a-judge a evaluation judge: name: doubao-seed-1-6-250615 diff --git a/config.yaml.simple b/config.yaml.simple index 95587a80..7c31c4dc 100644 --- a/config.yaml.simple +++ b/config.yaml.simple @@ -4,5 +4,6 @@ model: name: doubao-seed-1-6-250615 api_base: https://ark.cn-beijing.volces.com/api/v3/ api_key: - extra_headers: '{"x-is-encrypted": "true"}' - extra_body: '{"caching": {"type": "enabled"}, "thinking": {"type": "disabled"}, "expire_at": 1}' # expire after 1 hour \ No newline at end of file + encrypted: true # true | false + caching: enabled # enabled | disabled + max_llm_calls: 100 \ No newline at end of file diff --git a/veadk/agent.py b/veadk/agent.py index 768f2e8c..248b8559 100644 --- a/veadk/agent.py +++ b/veadk/agent.py @@ -47,29 +47,6 @@ logger = get_logger(__name__) -def _parse_user_model_extra_config() -> dict: - import json - - extra_headers = getenv("MODEL_AGENT_EXTRA_HEADERS", {}, allow_false_values=True) - extra_body = getenv("MODEL_AGENT_EXTRA_BODY", {}, allow_false_values=True) - - if extra_headers: - extra_headers = json.loads(extra_headers) - if extra_body: - extra_body = json.loads(extra_body) - if extra_body.get("expire_at", 0): - import time - - extra_body["expire_at"] = ( - int(time.time()) + int(extra_body["expire_at"]) * 3600 - ) - - return { - "extra_headers": extra_headers, - "extra_body": extra_body, - } - - class Agent(LlmAgent): """LLM-based Agent with Volcengine capabilities.""" @@ -97,7 +74,7 @@ class Agent(LlmAgent): model_api_key: str = Field(default_factory=lambda: getenv("MODEL_AGENT_API_KEY")) """The api key of the model for agent running.""" - model_extra_config: dict = Field(default_factory=_parse_user_model_extra_config) + model_extra_config: dict = Field(default_factory=dict) """The extra config to include in the model requests.""" tools: list[ToolUnion] = [] @@ -140,6 +117,8 @@ def model_post_init(self, __context: Any) -> None: "extra_body": body, } + logger.info(f"Model extra config: {self.model_extra_config}") + if not self.model: self.model = LiteLlm( model=f"{self.model_provider}/{self.model_name}", diff --git a/veadk/consts.py b/veadk/consts.py index fad5aecc..dd877690 100644 --- a/veadk/consts.py +++ b/veadk/consts.py @@ -14,6 +14,7 @@ import time +from veadk.config import getenv from veadk.version import VERSION DEFAULT_MODEL_AGENT_NAME = "doubao-seed-1-6-250615" @@ -21,13 +22,14 @@ DEFAULT_MODEL_AGENT_API_BASE = "https://ark.cn-beijing.volces.com/api/v3/" DEFAULT_MODEL_EXTRA_CONFIG = { "extra_headers": { - "x-is-encrypted": "true", + "x-is-encrypted": getenv("MODEL_AGENT_ENCRYPTED", "true"), "veadk-source": "veadk", "veadk-version": VERSION, }, "extra_body": { - "caching": {"type": "enabled"}, - "thinking": {"type": "disabled"}, + "caching": { + "type": getenv("MODEL_AGENT_CACHING", "enabled"), + }, "expire_at": int(time.time()) + 3600, # expire after 1 hour }, } diff --git a/veadk/runner.py b/veadk/runner.py index 54d8442a..ee67aeb8 100644 --- a/veadk/runner.py +++ b/veadk/runner.py @@ -27,12 +27,13 @@ from veadk.agents.loop_agent import LoopAgent from veadk.agents.parallel_agent import ParallelAgent from veadk.agents.sequential_agent import SequentialAgent +from veadk.config import getenv from veadk.evaluation import EvalSetRecorder +from veadk.integrations.ve_tos.ve_tos import VeTOS from veadk.memory.short_term_memory import ShortTermMemory from veadk.types import MediaMessage from veadk.utils.logger import get_logger from veadk.utils.misc import read_png_to_bytes -from veadk.integrations.ve_tos.ve_tos import VeTOS logger = get_logger(__name__) @@ -142,7 +143,13 @@ async def _run( if run_config is not None: stream_mode = run_config.streaming_mode else: - run_config = RunConfig(streaming_mode=stream_mode) + run_config = RunConfig( + streaming_mode=stream_mode, + max_llm_calls=int(getenv("MODEL_AGENT_MAX_LLM_CALLS", 100)), + ) + + logger.info(f"Run config: {run_config}") + try: async def event_generator(): @@ -231,7 +238,13 @@ async def run_with_raw_message( session_id: str, run_config: RunConfig | None = None, ): - run_config = RunConfig() if not run_config else run_config + run_config = ( + RunConfig(max_llm_calls=int(getenv("MODEL_AGENT_MAX_LLM_CALLS", 100))) + if not run_config + else run_config + ) + + logger.info(f"Run config: {run_config}") await self.short_term_memory.create_session( app_name=self.app_name, user_id=self.user_id, session_id=session_id