Handle KeyError for 'completion' in ClaudeAdapter

In this commit, we've added error handling for scenarios where the key 'completion' is not present in the 'claude_response' dictionary. This exception was causing the program to crash when the application tried to access the 'completion' key in the 'claude_response' dictionary.

Now, the 'get' method is used on the dictionary which allows the program to provide a default value of an empty string if the 'completion' key is not present. This prevents a KeyError from being raised and allows the program to continue execution. The affected methods are 'claude_to_chatgpt_response_stream' and 'claude_to_chatgpt_response'.

This is an important fix that improves the robustness of the code by handling potential exceptions arising from unexpected data structures.

claude_to_chatgpt/adapter.py

@@ -68,7 +68,7 @@ def openai_to_claude_params(self, openai_params):
         return claude_params
 
     def claude_to_chatgpt_response_stream(self, claude_response, prev_decoded_response):
-        completion_tokens = num_tokens_from_string(claude_response["completion"])
+        completion_tokens = num_tokens_from_string(claude_response.get("completion", ""))
         openai_response = {
             "id": f"chatcmpl-{str(time.time())}",
             "object": "chat.completion.chunk",
@@ -98,7 +98,7 @@ def claude_to_chatgpt_response_stream(self, claude_response, prev_decoded_respon
         return openai_response
 
     def claude_to_chatgpt_response(self, claude_response):
-        completion_tokens = num_tokens_from_string(claude_response["completion"])
+        completion_tokens = num_tokens_from_string(claude_response.get("completion", ""))
         openai_response = {
             "id": f"chatcmpl-{str(time.time())}",
             "object": "chat.completion",