Fix OpenAIFormatter to not rely on litellm imports

The litellm library doesn't export Delta, Choices, etc. directly.
Rewrote the formatter to build response dictionaries directly.

codai/models/parsers.py

 import time
 import uuid
-from litellm import ModelResponse, ChatCompletionChunk, Choices, StreamingChoices, Delta, Message, Usage
 
 
 class OpenAIFormatter:
@@ -25,24 +24,37 @@ class OpenAIFormatter:
             tool_calls: Optional list of tool calls to include
             
         Returns:
-            Dictionary representation of ModelResponse
+            Dictionary representation of the response
         """
-        return ModelResponse(
-            id=self.id,
-            model=self.model_name,
-            object="chat.completion",
-            created=int(time.time()),
-            choices=[Choices(
-                finish_reason="tool_calls" if tool_calls else "stop",
-                index=0,
-                message=Message(content=text if not tool_calls else None, role="assistant", tool_calls=tool_calls)
-            )],
-            usage=Usage(
-                prompt_tokens=prompt_tokens,
-                completion_tokens=completion_tokens,
-                total_tokens=prompt_tokens + completion_tokens
-            )
-        ).model_dump()
+        message = {
+            "role": "assistant",
+            "content": text if not tool_calls else None,
+        }
+        if tool_calls:
+            message["tool_calls"] = tool_calls
+        
+        choice = {
+            "index": 0,
+            "message": message,
+            "finish_reason": "tool_calls" if tool_calls else "stop",
+        }
+        
+        return {
+            "id": self.id,
+            "object": "chat.completion",
+            "created": int(time.time()),
+            "model": self.model_name,
+            "choices": [choice],
+            "usage": {
+                "prompt_tokens": prompt_tokens,
+                "completion_tokens": completion_tokens,
+                "total_tokens": prompt_tokens + completion_tokens,
+            },
+            "provider": {
+                "provider_name": "coderai",
+                "provider_id": "coderai",
+            },
+        }
     
     def format_chunk(self, delta_text: str, is_final: bool = False, usage: dict = None) -> dict:
         """Format a streaming chunk response.
@@ -53,20 +65,31 @@ class OpenAIFormatter:
             usage: Optional usage information (typically only sent on final chunk)
             
         Returns:
-            Dictionary representation of ChatCompletionChunk
+            Dictionary representation of the chunk
         """
-        return ChatCompletionChunk(
-            id=self.id,
-            model=self.model_name,
-            object="chat.completion.chunk",
-            created=int(time.time()),
-            choices=[StreamingChoices(
-                finish_reason="stop" if is_final else None,
-                index=0,
-                delta=Delta(content=delta_text, role="assistant")
-            )],
-            usage=usage  # Only send usage on the final chunk
-        ).model_dump()
+        delta = {
+            "content": delta_text,
+            "role": "assistant",
+        }
+        
+        choice = {
+            "index": 0,
+            "delta": delta,
+            "finish_reason": "stop" if is_final else None,
+        }
+        
+        chunk = {
+            "id": self.id,
+            "object": "chat.completion.chunk",
+            "created": int(time.time()),
+            "model": self.model_name,
+            "choices": [choice],
+        }
+        
+        if usage and is_final:
+            chunk["usage"] = usage
+            
+        return chunk
     
     def format_final_chunk(self, usage: dict = None) -> dict:
         """Format the final streaming chunk with usage information.
@@ -75,17 +98,28 @@ class OpenAIFormatter:
             usage: Usage statistics dictionary with prompt_tokens, completion_tokens, total_tokens
             
         Returns:
-            Dictionary representation of the final ChatCompletionChunk
+            Dictionary representation of the final chunk
         """
-        return ChatCompletionChunk(
-            id=self.id,
-            model=self.model_name,
-            object="chat.completion.chunk",
-            created=int(time.time()),
-            choices=[StreamingChoices(
-                finish_reason="stop",
-                index=0,
-                delta=Delta(content=None, role="assistant")
-            )],
-            usage=usage
-        ).model_dump()
+        delta = {
+            "content": None,
+            "role": "assistant",
+        }
+        
+        choice = {
+            "index": 0,
+            "delta": delta,
+            "finish_reason": "stop",
+        }
+        
+        chunk = {
+            "id": self.id,
+            "object": "chat.completion.chunk",
+            "created": int(time.time()),
+            "model": self.model_name,
+            "choices": [choice],
+        }
+        
+        if usage:
+            chunk["usage"] = usage
+            
+        return chunk