New response.excute_tool_calls(), refs #1007

llm/models.py

@@ -192,6 +192,10 @@ class ToolResult:
     tool_call_id: Optional[str] = None
 
 
+class CancelToolCall(Exception):
+    pass
+
+
 @dataclass
 class Prompt:
     _prompt: Optional[str]
@@ -324,7 +328,7 @@ class Conversation(_BaseConversation):
         tool_results: Optional[List[ToolResult]] = None,
         details: bool = False,
         key: Optional[str] = None,
-        **options,
+        options: Optional[dict] = None,
     ):
         self.model._validate_attachments(attachments)
         return ChainResponse(
@@ -338,7 +342,7 @@ class Conversation(_BaseConversation):
                 tool_results=tool_results,
                 system_fragments=system_fragments,
                 model=self.model,
-                options=self.model.Options(**options),
+                options=self.model.Options(**(options or {})),
             ),
             model=self.model,
             stream=stream,
@@ -482,6 +486,42 @@ class _BaseResponse:
     def add_tool_call(self, tool_call: ToolCall):
         self._tool_calls.append(tool_call)
 
+    def execute_tool_calls(
+        self,
+        *,
+        before_call: Optional[Callable[[Tool, ToolCall], None]] = None,
+        after_call: Optional[Callable[[Tool, ToolCall, ToolResult], None]] = None,
+    ) -> List[ToolResult]:
+        tool_results = []
+        tools_by_name = {tool.name: tool for tool in self.prompt.tools}
+        # TODO: make this work async
+        for tool_call in self.tool_calls():
+            tool = tools_by_name.get(tool_call.name)
+            if tool is None:
+                raise CancelToolCall("Unknown tool: {}".format(tool_call.name))
+            if before_call:
+                # This may raise CancelToolCall:
+                before_call(tool, tool_call)
+            if not tool.implementation:
+                raise CancelToolCall(
+                    "No implementation available for tool: {}".format(tool_call.name)
+                )
+            try:
+                result = tool.implementation(**tool_call.arguments)
+                if not isinstance(result, str):
+                    result = json.dumps(result, default=repr)
+            except Exception as ex:
+                result = f"Error: {ex}"
+            tool_result = ToolResult(
+                name=tool_call.name,
+                output=result,
+                tool_call_id=tool_call.tool_call_id,
+            )
+            if after_call:
+                after_call(tool, tool_call, tool_result)
+            tool_results.append(tool_result)
+        return tool_results
+
     def set_usage(
         self,
         *,
@@ -1023,7 +1063,7 @@ class _BaseChainResponse:
         conversation: _BaseConversation,
         key: Optional[str] = None,
         details: bool = False,
-        chain_limit: int = 5,
+        chain_limit: int = 10,
     ):
         self.prompt = prompt
         self.model = model
@@ -1055,44 +1095,8 @@ class _BaseChainResponse:
             self._responses.append(response)
             if count > self.chain_limit:
                 raise ValueError(f"Chain limit of {self.chain_limit} exceeded. ")
-            tool_calls = response.tool_calls()
-            if not tool_calls:
-                return
-            tools_by_name = {
-                tool.name: tool.implementation for tool in response.prompt.tools
-            }
-            tool_results = []
-            for tool_call in tool_calls:
-                if details:
-                    yield "\nTool call requested: {}({})\n".format(
-                        tool_call.name,
-                        ", ".join(
-                            f"{k}={repr(v)}" for k, v in tool_call.arguments.items()
-                        ),
-                    )
-                implementation = tools_by_name.get(tool_calls[0].name)
-                if not implementation:
-                    # TODO: send this as an error instead?
-                    raise ValueError(f"Tool {tool_call.name} not found")
-                try:
-                    result = implementation(**tool_call.arguments)
-                    if not isinstance(result, str):
-                        result = json.dumps(result, default=repr)
-                    if details:
-                        yield f"\n{result}\n"
-                except Exception as ex:
-                    if details:
-                        yield f"\nError:\ns{ex}\n"
-                    result = f"Error: {ex}"
-                tool_results.append(
-                    ToolResult(
-                        name=tool_call.name,
-                        output=result,
-                        tool_call_id=tool_call.tool_call_id,
-                    )
-                )
-            if not tool_results:
-                break
+            # This could raise llm.CancelToolCall:
+            tool_results = response.execute_tool_calls()
             response = Response(
                 Prompt(
                     "",
@@ -1256,7 +1260,7 @@ class _Model(_BaseModel):
         tools: Optional[List[Tool]] = None,
         tool_results: Optional[List[ToolResult]] = None,
         details: bool = False,
-        **options,
+        options: Optional[dict] = None,
     ):
         return self.conversation().chain(
             prompt=prompt,
@@ -1269,7 +1273,7 @@ class _Model(_BaseModel):
             tools=tools,
             tool_results=tool_results,
             details=details,
-            **options,
+            options=options,
         )