Modify max_completion_tokens parameter for tool call accuracy evaluator (#42544)

* support 5 levels, evaluate all tools at once

* Update sample notebook and change log

* Add missing import

* Modify test cases to match the new output format

* Modify other test file to match the new output format

* Fixed parsing of results

* Change key name in output

* Spell check fixes

* Minor prompt update

* Update result key to tool_call_accuracy

* Delete test_new_evaluator.ipynb

* Added field names and messages as constants

* Additional note in prompt

* Re-add the temperature to the prompty file

* Removed 'applicable' field and print statement

* Move excess/missing tool calls fields under additional details

* Typo fix and removal of redundant field in the prompt

* Modify per_tool_call_details field's name to details

* Revert "Modify per_tool_call_details field's name to details"

This reverts commit 2c3ce50.

* Revert 'Merge branch 'main' into selshafey/improve_tool_call_accuracy'

* Black reformat

* Reformat with black

* To re-trigger build pipelines

* Add notebook for bugbash

* modify bugbash notebook

* Add support for built-in tools for Tool Call Accuracy Evaluator

* Remove bugbash notebook

* Resolve issues with merge

* Fix id value

* Use existing built-in tool definitions

* Run black

* Prompt modifications

* Add test cases for built-in tools

* Handle converter format

* Add test cases for converter format

* Support only converter format

* Revert tool definitions to be required, run black

* Change max_completion_tokens to max_tokens

* merged with main

* re-merge with main

* run black

---------

Co-authored-by: Salma Elshafey <[email protected]>

Author: salma-elshafey

sdk/evaluation/azure-ai-evaluation/azure/ai/evaluation/_evaluators/_tool_call_accuracy/tool_call_accuracy.prompty

@@ -4,7 +4,8 @@ description: Evaluates Tool Call Accuracy for tool used by agent
 model:
   api: chat
   parameters:
-    max_completion_tokens: 3000
+    temperature: 0.0
+    max_tokens: 3000
     top_p: 1.0
     presence_penalty: 0
     frequency_penalty: 0