Cleanup configs

dockers/llm.rag.service/.env-local

@@ -0,0 +1,9 @@
+WEAVIATE_URI_WITH_PORT=localhost:8080
+WEAVIATE_GRPC_URI_WITH_PORT=localhost:50051
+WEAVIATE_INDEX_NAME=my_custom_index
+
+MODEL_LLM_SERVER_URL=http://localhost:9000/v1
+
+SQL_SEARCH_DB_AND_MODEL_PATH=/tmp/db/
+
+RAGLLM_LOGS_PATH=logs/ragllm.log

dockers/llm.rag.service/.gitignore

@@ -0,0 +1,2 @@
+.env
+uv.lock

dockers/llm.rag.service/Dockerfile

@@ -47,6 +47,7 @@ COPY serveragllm.py .
 COPY logging_config.py .
 COPY serverragllm_csv_to_weaviate_local.py .
 COPY common.py .
+COPY config.py .
 COPY pyproject.toml .
 
 EXPOSE 8000

dockers/llm.rag.service/common.py

@@ -478,16 +478,13 @@ def get_answer_with_settings_with_weaviate_filter(
     question,
     vectorstore,
     client,
-    model_id,
-    max_tokens,
-    model_temperature,
     system_prompt,
     relevant_docs,
-    llm_server_url,
     sql_search_db_and_model_path,
     alpha,
     max_context_length,
     sql_ticket_source,
+    llm_settings,
 ):
 
     search_type = question_router(question, sql_search_db_and_model_path)
@@ -499,10 +496,10 @@ def get_answer_with_settings_with_weaviate_filter(
 
             return get_sql_answer(
                 question,
-                model_id,
-                max_tokens,
-                model_temperature,
-                llm_server_url,
+                llm_settings.model_id,
+                llm_settings.max_tokens,
+                llm_settings.model_temperature,
+                llm_settings.llm_server_url,
                 sql_search_db_and_model_path,
                 max_context_length,
                 sql_ticket_source,
@@ -530,9 +527,9 @@ def get_answer_with_settings_with_weaviate_filter(
                 question,
                 retriever,
                 client,
-                model_id,
-                max_tokens,
-                model_temperature,
+                llm_settings.model_id,
+                llm_settings.max_tokens,
+                llm_settings.model_temperature,
                 system_prompt,
             )
 

dockers/llm.rag.service/config.py

@@ -0,0 +1,168 @@
+from typing import Optional
+
+from pydantic import ConfigDict, Field, field_validator
+from pydantic_settings import BaseSettings
+
+
+SYSTEM_PROMPT_DEFAULT = """You are a specialized support ticket assistant. Format your responses following these rules:
+                1. Answer the provided question only using the provided context.
+                2. Do not add the provided context to the generated answer.
+                3. Include relevant technical details when present or provide a summary of the comments in the ticket.
+                4. Include the submitter, assignee and collaborator for a ticket when this info is available.
+                5. If the question cannot be answered with the given context, please say so and do not attempt to provide an answer.
+                6. Do not create new questions related to the given question, instead answer only the provided question.
+                7. Provide a clear, direct and factual answer.
+                """
+
+
+def validate_float(value):
+    if type(value) == float:
+        return value
+    try:
+        return float(value.strip("'").strip('"'))
+    except (TypeError, ValueError):
+        raise ValueError("Value must be convertible to a float")
+
+
+def validate_int(value):
+    if type(value) == int:
+        return value
+    try:
+        return int(value.strip("'").strip("\""))
+    except (TypeError, ValueError):
+        raise ValueError("Value must be convertible to an integer")
+
+
+class WeaviateSettings(BaseSettings):
+    model_config = ConfigDict(env_file=".env", extra="ignore")
+
+    weaviate_uri: Optional[str] = Field(
+        None,
+        alias="WEAVIATE_URI_WITH_PORT",
+    )
+    weaviate_grpc_uri: Optional[str] = Field(
+        None,
+        alias="WEAVIATE_GRPC_URI_WITH_PORT",
+    )
+    weaviate_index_name: Optional[str] = Field(
+        None,
+        alias="WEAVIATE_INDEX_NAME",
+    )
+    embedding_model_name: Optional[str] = Field(
+        default="sentence-transformers/all-MiniLM-L6-v2",
+        alias="EMBEDDING_MODEL_NAME",
+    )
+
+    def is_set(self) -> bool:
+        return all(
+            [self.weaviate_uri, self.weaviate_grpc_uri, self.weaviate_index_name]
+        )
+
+    def get_weaviate_uri(self):
+        return self.weaviate_uri.split(":")[0]
+
+    def get_weaviate_port(self):
+        return int(self.weaviate_uri.split(":")[1])
+
+    def get_weaviate_grpc_uri(self):
+        return self.weaviate_grpc_uri.split(":")[0]
+
+    def get_weaviate_grpc_port(self):
+        return int(self.weaviate_grpc_uri.split(":")[1])
+
+
+class LlmSettings(BaseSettings):
+    model_config = ConfigDict(env_file=".env", extra="ignore")
+
+    llm_server_url: Optional[str] = Field(
+        None,
+        alias="MODEL_LLM_SERVER_URL",
+    )
+    model_id: Optional[str] = Field(
+        default="rubra-ai/Phi-3-mini-128k-instruct",
+        alias="MODEL_ID",
+    )
+    max_tokens: Optional[int] = Field(
+        default=256,
+        alias="MAX_TOKENS",
+    )
+    model_temperature: Optional[float] = Field(
+        default=0.01,
+        alias="MODEL_TEMPERATURE",
+    )
+
+    @field_validator("llm_server_url")
+    @classmethod
+    def ensure_v1_endpoint(cls, v: str) -> str:
+        if not v.endswith("/v1"):
+            v = v + "/v1"
+        return v
+
+    @field_validator("max_tokens", mode="before")
+    @classmethod
+    def validate_max_tokens(cls, v):
+        return validate_int(v)
+
+    @field_validator("model_temperature", mode="before")
+    @classmethod
+    def validate_model_temperature(cls, v):
+        return validate_float(v)
+
+
+class HybridSearchSettings(BaseSettings):
+    model_config = ConfigDict(env_file=".env", extra="ignore")
+
+    alpha: Optional[float] = Field(
+        default=0.5,
+        alias="WEAVIATE_HYBRID_ALPHA",
+    )
+    relevant_docs: Optional[int] = Field(
+        default=2,
+        alias="RELEVANT_DOCS",
+    )
+    system_prompt: Optional[str] = Field(
+        default=SYSTEM_PROMPT_DEFAULT,
+        alias="SYSTEM_PROMPT",
+    )
+
+    @field_validator("alpha", mode="before")
+    @classmethod
+    def validate_alpha(cls, v):
+        return validate_float(v)
+
+    @field_validator("relevant_docs", mode="before")
+    @classmethod
+    def validate_relevant_docs(cls, v):
+        return validate_int(v)
+
+
+class SqlSearchSettings(BaseSettings):
+    model_config = ConfigDict(env_file=".env", extra="ignore")
+
+    db_and_model_path: Optional[str] = Field(
+        default="/app/db/",
+        alias="SQL_SEARCH_DB_AND_MODEL_PATH",
+    )
+    max_context_length: Optional[int] = Field(
+        default=8192,
+        alias="MODEL_MAX_CONTEXT_LEN",
+    )
+    # TODO: Read sources from DB instead of using this
+    ticket_source: Optional[str] = Field(
+        default="https://zendesk.com/api/v2/tickets/",
+        alias="SQL_TICKET_SOURCE",
+    )
+
+    @field_validator("max_context_length", mode="before")
+    @classmethod
+    def validate_max_context_length(cls, v):
+        return validate_int(v)
+
+
+class LoggingSettings(BaseSettings):
+    model_config = ConfigDict(env_file=".env", extra="ignore")
+
+    log_file_path: Optional[str] = Field(
+        default="/app/logs/ragllm.log",
+        alias="RAGLLM_LOGS_PATH",
+    )

dockers/llm.rag.service/logging_config.py

@@ -2,18 +2,19 @@
 import os
 from logging.handlers import TimedRotatingFileHandler
 
-# When running locally: export RAGLLM_LOGS_PATH=logs/ragllm.log
-log_file_path = os.getenv("RAGLLM_LOGS_PATH") or "/app/logs/ragllm.log"
+from config import LoggingSettings
+
+settings = LoggingSettings()
 os.makedirs(
-    os.path.dirname(log_file_path), exist_ok=True
+    os.path.dirname(settings.log_file_path), exist_ok=True
 )  # Ensure log directory exists
 
 logging.basicConfig(
     level=logging.INFO,
     format="%(asctime)s - %(levelname)s - %(filename)s:%(lineno)d -- %(message)s",
     handlers=[
         # Log to file, rotate every 1H and store files from last 24 hrs * 7 days files == 168H data
-        TimedRotatingFileHandler(log_file_path, when="h", interval=1, backupCount=168),
+        TimedRotatingFileHandler(settings.log_file_path, when="h", interval=1, backupCount=168),
         logging.StreamHandler(),  # Also log to console
     ],
 )

dockers/llm.rag.service/proxy_app.py

@@ -2,17 +2,17 @@
 
 import uvicorn
 
+from config import WeaviateSettings
+
 
 def main():
     # Check environment variables to determine which app to run
-    weaviate_url = os.getenv("WEAVIATE_URI_WITH_PORT")
-    weaviate_grpc_url = os.getenv("WEAVIATE_GRPC_URI_WITH_PORT")
-    weaviate_index = os.getenv("WEAVIATE_INDEX_NAME")
+    weaviate_settings = WeaviateSettings()
 
     host = os.environ.get("APP_HOST", "0.0.0.0")
     port = int(os.environ.get("APP_PORT", "8000"))
 
-    if weaviate_url and weaviate_grpc_url and weaviate_index:
+    if weaviate_settings.is_set():
         # Run the CSV to Weaviate app
         print(f"Starting Weaviate app on {host}:{port}")
         uvicorn.run("serverragllm_csv_to_weaviate_local:app", host=host, port=port)