Merge branch 'mrh-online-dev' of github.com:BRUNIX-AI/assistance-engine into mrh-online-dev

Docker/config.py

@@ -1,81 +0,0 @@
-from pathlib import Path
-from pydantic_settings import BaseSettings, SettingsConfigDict
-from dotenv import load_dotenv
-from datetime import timedelta
-import warnings
-import os
-
-load_dotenv()
-
-MODEL_NAME=os.getenv("MODEL_NAME", "gpt-3.5-turbo")
-OPENAI_API_KEY=os.getenv("OPENAI_API_KEY", "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")
- 
-OLLAMA_URL=os.getenv("OLLAMA_URL", "http://host.docker.internal:11434")
-OLLAMA_LOCAL_URL=os.getenv("OLLAMA_LOCAL_URL", "http://localhost:11434")
-OLLAMA_MODEL_NAME=os.getenv("OLLAMA_MODEL_NAME", "qwen3-0.6B:latest")
-OLLAMA_EMB_MODEL_NAME=os.getenv("OLLAMA_EMB_MODEL_NAME", "qwen3-0.6B-emb:latest")
- 
-LANGFUSE_HOST=os.getenv("LANGFUSE_HOST", "http://45.77.119.180")
-LANGFUSE_PUBLIC_KEY=os.getenv("LANGFUSE_PUBLIC_KEY", "pk-lf-0e6db694-3e95-4dd4-aedf-5a2694267058")
-LANGFUSE_SECRET_KEY=os.getenv("LANGFUSE_SECRET_KEY", "sk-lf-dbf28bb9-15bb-4d03-a8c3-05caa3e3905f")
- 
-ELASTICSEARCH_URL=os.getenv("ELASTICSEARCH_URL", "http://host.docker.internal:9200")
-ELASTICSEARCH_LOCAL_URL=os.getenv("ELASTICSEARCH_LOCAL_URL", "http://localhost:9200")
-ELASTICSEARCH_INDEX=os.getenv("ELASTICSEARCH_INDEX", "avap-docs-test")
- 
-DATABASE_URL=os.getenv("DATABASE_URL", "postgresql://postgres:brunix_pass@host.docker.internal:5432/postgres")
- 
-KUBECONFIG_PATH=os.getenv("KUBECONFIG_PATH", "kubernetes/kubeconfig.yaml")
- 
-HF_TOKEN=os.getenv("HF_TOKEN", "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")
-HF_EMB_MODEL_NAME=os.getenv("HF_EMB_MODEL_NAME", "Qwen/Qwen3-Embedding-0.6B")
-
-MRH_AVAP_DATA_PATH_=os.getenv("MRH_AVAP_DATA_PATH_", "/home/pseco/VsCodeProjects/assistance-engine/data/")
-
-MRH_AVAP_MODELS_PATH_=os.getenv("MRH_AVAP_MODELS_PATH_", "/home/pseco/VsCodeProjects/assistance-engine/data/models")
-MRH_AVAP_RAW_PATH_=os.getenv("MRH_AVAP_RAW_PATH_", "/home/pseco/VsCodeProjects/assistance-engine/data/raw")
-MRH_AVAP_PROCESSED_PATH_=os.getenv("MRH_AVAP_PROCESSED_PATH_", "/home/pseco/VsCodeProjects/assistance-engine/data/processed")
-MRH_AVAP_INTERIM_PATH_=os.getenv("MRH_AVAP_INTERIM_PATH_", "/home/pseco/VsCodeProjects/assistance-engine/data/interim")
-MRH_AVAP_EXTERNAL_PATH_=os.getenv("MRH_AVAP_EXTERNAL_PATH_", "/home/pseco/VsCodeProjects/assistance-engine/data/external")
-
-
-class Settings(BaseSettings):
-    raw_path_: str
-    processed_path_: str
-    models_path_:str
-    interim_path_:str
-    external_path_:str
-
-    model_config = SettingsConfigDict(
-        env_prefix="mrh_avap_",
-        env_file=".env",
-        env_file_encoding="utf-8",
-        case_sensitive=False,
-        extra="ignore",
-    )
-
-    @property
-    def raw_path(self) -> Path:
-        return Path(self.raw_path_)
-
-    @property
-    def processed_path(self) -> Path:
-        return Path(self.processed_path_)
-
-    @property
-    def proj_root(self) -> Path:
-        return Path(__file__).resolve().parents[1]
-
-    @property
-    def interim_path(self) -> Path:
-        return Path(self.interim_path_)
-    
-    @property
-    def external_path(self) -> Path:
-        return Path(self.external_path_)
-
-    @property
-    def models_path(self) -> Path:
-        return Path(self.models_path_)
-
-settings = Settings()

Docker/docker-compose.yaml

@@ -13,7 +13,7 @@ services:
       LANGFUSE_HOST: ${LANGFUSE_HOST}
       LANGFUSE_PUBLIC_KEY: ${LANGFUSE_PUBLIC_KEY}
       LANGFUSE_SECRET_KEY: ${LANGFUSE_SECRET_KEY}
-      OLLAMA_URL: ${OLLAMA_LOCAL_URL}
+      OLLAMA_URL: ${OLLAMA_URL}
       OLLAMA_MODEL_NAME: ${OLLAMA_MODEL_NAME}
       OLLAMA_EMB_MODEL_NAME: ${OLLAMA_EMB_MODEL_NAME}
     

Docker/src/instances.py

@@ -1,24 +0,0 @@
-import os
-
-from langchain_elasticsearch import ElasticsearchStore
-
-from utils.llm_factory import create_chat_model
-from utils.emb_factory import create_embedding_model
-
-llm = create_chat_model(
-    provider="ollama",
-    model=os.getenv("OLLAMA_MODEL_NAME"),
-    temperature=0,
-    validate_model_on_init=True,
-)
-embeddings = create_embedding_model(
-    provider="ollama",
-    model=os.getenv("OLLAMA_EMB_MODEL_NAME"),
-)
-vector_store = ElasticsearchStore(
-    es_url=os.getenv("ELASTICSEARCH_URL"),
-    index_name=os.getenv("ELASTICSEARCH_INDEX"),
-    embedding=embeddings,
-    query_field="text",
-    vector_query_field="vector",
-)

Docker/src/server.py

@@ -1,46 +1,34 @@
 import logging
 import os
-import sys
 from concurrent import futures
-from pathlib import Path
-from typing import Any
 
 import brunix_pb2
 import brunix_pb2_grpc
 import grpc
 from grpc_reflection.v1alpha import reflection
-from langchain_core.prompts import ChatPromptTemplate
 from langchain_elasticsearch import ElasticsearchStore
 
 from utils.llm_factory import create_chat_model
 from utils.emb_factory import create_embedding_model
 from graph import build_graph
 
-# PROJECT_ROOT = Path(__file__).resolve().parents[2]
-# if str(PROJECT_ROOT) not in sys.path:
-#     sys.path.insert(0, str(PROJECT_ROOT))
 
 logging.basicConfig(level=logging.INFO)
 logger = logging.getLogger("brunix-engine")
 
-
-def _provider_kwargs(provider: str, base_url: str) -> dict[str, Any]:
-    if provider == "ollama":
-        return {"base_url": base_url}
-    return {}
-
-
 class BrunixEngine(brunix_pb2_grpc.AssistanceEngineServicer):
     def __init__(self):
         self.llm = create_chat_model(
             provider="ollama",
             model=os.getenv("OLLAMA_MODEL_NAME"),
+            base_url=os.getenv("OLLAMA_URL"),
             temperature=0,
             validate_model_on_init=True,
         )
         self.embeddings = create_embedding_model(
             provider="ollama",
             model=os.getenv("OLLAMA_EMB_MODEL_NAME"),
+            base_url=os.getenv("OLLAMA_URL"),
         )
         self.vector_store = ElasticsearchStore(
             es_url=os.getenv("ELASTICSEARCH_URL"),