Update readme for separate llm engine executables

2025-05-10 17:09:29 +02:00
parent f8365ef698
commit 895a533e01
5 changed files with 297 additions and 233 deletions
--- a/sia/base_agent.py
+++ b/sia/base_agent.py
@@ -1,77 +1,74 @@
-from abc import ABC
-import xml.etree.ElementTree as ET
-
-from .llm_engine import LlmEngine
-from .response_parser import ResponseParser
-from .system_metrics import SystemMetrics
-from .util import pretty_print_element
-from .working_memory import WorkingMemory
-from .xml_validator import XMLValidator
-
-class BaseAgent(ABC):
-    """
-    Abstract base class for SIA agents.
-    
-    Provides core functionality for maintaining working memory, system metrics,
-    and coordinating components for LLM inference.
-    """
-    
-    def __init__(
-            self,
-            system_prompt: str,
-            action_schema: str, 
-            working_memory: WorkingMemory,
-            metrics: SystemMetrics,
-            validator: XMLValidator,
-            parser: ResponseParser,
-        ):
-        """
-        Initialize agent with required components.
-        """
-        self._system_prompt = system_prompt
-        self._action_schema = action_schema
-        self._working_memory = working_memory
-        self._metrics = metrics
-        self._validator = validator
-        self._parser = parser
-    
-    @property
-    def system_prompt(self) -> str:
-        """Get the system prompt."""
-        return f"{self._system_prompt}\n{self._action_schema}"
-            
-    def _compile_context(self, llmEngine: LlmEngine) -> str:
-        """
-        Compile the current context for LLM inference.
-        Includes system metrics and working memory entries.
-        
-        Returns:
-            str: Complete context as XML string
-        """
-        memory_context = self._working_memory.generate_context()
-        metrics_data = self._metrics.get_metrics()
-        
-        # Create context element
-        context = ET.Element("context")
-        context.set("time", metrics_data["timestamp"])
-        context.set("memory_used", str(metrics_data["memory_used"]))
-        context.set("memory_total", str(metrics_data["memory_total"]))
-        context.set("disk_used", str(metrics_data["disk_used"]))
-        context.set("disk_total", str(metrics_data["disk_total"]))
-        context.set("stdin", str(self._parser.io_buffer.buffer_length()))
-        context.set("context", "100%")
-
-        for entry in memory_context:
-            context.append(entry)
-
-        context_str = pretty_print_element(context)
-        
-        # Calculate token usage percentage
-        token_count = llmEngine.token_count(self.system_prompt, context_str)
-        token_limit = llmEngine.token_limit()
-        context_usage = (float(token_count) / float(token_limit)) * 100.0
-        
-        # Update context usage metric
-        context.set("context", f"{str(round(context_usage, 2))}%")
-        
+from abc import ABC
+import xml.etree.ElementTree as ET
+
+from .llm_engine import LlmEngine
+from .response_parser import ResponseParser
+from .system_metrics import SystemMetrics
+from .util import pretty_print_element
+from .working_memory import WorkingMemory
+
+class BaseAgent(ABC):
+    """
+    Abstract base class for SIA agents.
+    
+    Provides core functionality for maintaining working memory, system metrics,
+    and coordinating components for LLM inference.
+    """
+    
+    def __init__(
+            self,
+            system_prompt: str,
+            action_schema: str, 
+            working_memory: WorkingMemory,
+            metrics: SystemMetrics,
+            parser: ResponseParser,
+        ):
+        """
+        Initialize agent with required components.
+        """
+        self._system_prompt = system_prompt
+        self._action_schema = action_schema
+        self._working_memory = working_memory
+        self._metrics = metrics
+        self._parser = parser
+    
+    @property
+    def system_prompt(self) -> str:
+        """Get the system prompt."""
+        return f"{self._system_prompt}\n{self._action_schema}"
+            
+    def _compile_context(self, llmEngine: LlmEngine) -> str:
+        """
+        Compile the current context for LLM inference.
+        Includes system metrics and working memory entries.
+        
+        Returns:
+            str: Complete context as XML string
+        """
+        memory_context = self._working_memory.generate_context()
+        metrics_data = self._metrics.get_metrics()
+        
+        # Create context element
+        context = ET.Element("context")
+        context.set("time", metrics_data["timestamp"])
+        context.set("memory_used", str(metrics_data["memory_used"]))
+        context.set("memory_total", str(metrics_data["memory_total"]))
+        context.set("disk_used", str(metrics_data["disk_used"]))
+        context.set("disk_total", str(metrics_data["disk_total"]))
+        context.set("stdin", str(self._parser.io_buffer.buffer_length()))
+        context.set("context", "100%")
+
+        for entry in memory_context:
+            context.append(entry)
+
+        context_str = pretty_print_element(context)
+        
+        # Calculate token usage percentage
+        token_count = llmEngine.token_count(self.system_prompt, context_str)
+        token_limit = llmEngine.token_limit()
+        context_usage = (float(token_count) / float(token_limit)) * 100.0
+        
+        # Update context usage metric
+        context.set("context", f"{str(round(context_usage, 2))}%")
+        
        return pretty_print_element(context)