Fixed arguments passed to InvalidTool.run(). (#1340)

[InvalidTool.run()](https://github.com/hwchase17/langchain/blob/aa7e6ae88e1aa6777bc018fcb7f185bb6f1b6f21/langchain/agents/tools.py#L43)
returns "{arg}is not a valid tool, try another one.".
However, no function name is actually given in the argument.
This causes LLM to be stuck in a loop, unable to find the right tool.

This may resolve these Issues.
langchain-ai/langchain#998
langchain-ai/langchain#702

langchain/agents/agent.py

@@ -426,7 +426,7 @@ def _take_next_step(
             )
         else:
             observation = InvalidTool().run(
-                output.tool_input,
+                output.tool,
                 verbose=self.verbose,
                 color=None,
                 llm_prefix="",
@@ -473,7 +473,7 @@ async def _atake_next_step(
             )
         else:
             observation = await InvalidTool().arun(
-                output.tool_input,
+                output.tool,
                 verbose=self.verbose,
                 color=None,
                 llm_prefix="",