Adding token usage support

OAI/types/chat_completion.py

@@ -32,8 +32,6 @@ class ChatCompletionResponse(BaseModel):
     created: int = Field(default_factory=lambda: int(time()))
     model: str
     object: str = "chat.completion"
-
-    # TODO: Add usage stats
     usage: Optional[UsageStats] = None
 
 class ChatCompletionStreamChunk(BaseModel):

OAI/types/common.py

@@ -8,8 +8,8 @@ class LogProbs(BaseModel):
     top_logprobs: List[Dict[str, float]] = Field(default_factory=list)
 
 class UsageStats(BaseModel):
-    completion_tokens: int
     prompt_tokens: int
+    completion_tokens: int
     total_tokens: int
 
 class CommonCompletionRequest(BaseModel):

OAI/types/completion.py

@@ -22,6 +22,4 @@ class CompletionResponse(BaseModel):
     created: int = Field(default_factory=lambda: int(time()))
     model: str
     object: str = "text_completion"
-
-    # TODO: Add usage stats
     usage: Optional[UsageStats] = None