From 800b6010c0bf76aadf678bc38a507b749fb9774c Mon Sep 17 00:00:00 2001
From: AllentDan <41138331+AllentDan@users.noreply.github.com>
Date: Thu, 23 Jan 2025 14:25:40 +0800
Subject: [PATCH] More arguments in api_client, update docstrings (#3077)

---
 lmdeploy/serve/openai/api_client.py | 31 +++++++++++++++++++++++++----
 lmdeploy/serve/openai/api_server.py |  4 ++++
 2 files changed, 31 insertions(+), 4 deletions(-)

diff --git a/lmdeploy/serve/openai/api_client.py b/lmdeploy/serve/openai/api_client.py
index 03b1ef708..79fd04570 100644
--- a/lmdeploy/serve/openai/api_client.py
+++ b/lmdeploy/serve/openai/api_client.py
@@ -103,9 +103,14 @@ def chat_completions_v1(self,
                             frequency_penalty: Optional[float] = 0.0,
                             user: Optional[str] = None,
                             repetition_penalty: Optional[float] = 1.0,
-                            session_id: Optional[int] = -1,
                             ignore_eos: Optional[bool] = False,
                             skip_special_tokens: Optional[bool] = True,
+                            spaces_between_special_tokens: Optional[bool] = True,
+                            top_k: int = 40,
+                            min_new_tokens: Optional[int] = None,
+                            min_p: float = 0.0,
+                            logit_bias: Optional[Dict[str, float]] = None,
+                            stream_options: Optional[Dict] = None,
                             **kwargs):
         """Chat completion v1.
 
@@ -128,7 +133,20 @@ def chat_completions_v1(self,
             ignore_eos (bool): indicator for ignoring eos
             skip_special_tokens (bool): Whether or not to remove special tokens
                 in the decoding. Default to be True.
-            session_id (int): Deprecated.
+            spaces_between_special_tokens (bool): Whether or not to add spaces
+                around special tokens. The behavior of Fast tokenizers is to have
+                this to False. This is setup to True in slow tokenizers.
+            top_k (int): The number of the highest probability vocabulary
+                tokens to keep for top-k-filtering
+            min_new_tokens (int): To generate at least numbers of tokens.
+            min_p (float): Minimum token probability, which will be scaled by the
+                probability of the most likely token. It must be a value between
+                0 and 1. Typical values are in the 0.01-0.2 range, comparably
+                selective as setting `top_p` in the 0.99-0.8 range (use the
+                opposite of normal `top_p` values)
+            logit_bias (Dict): Bias to logits. Only supported in pytorch engine.
+            stream_options: Options for streaming response. Only set this when you
+                set stream: true.
 
         Yields:
             json objects in openai formats
@@ -229,9 +247,10 @@ def completions_v1(
             user: Optional[str] = None,
             # additional argument of lmdeploy
             repetition_penalty: Optional[float] = 1.0,
-            session_id: Optional[int] = -1,
             ignore_eos: Optional[bool] = False,
             skip_special_tokens: Optional[bool] = True,
+            spaces_between_special_tokens: Optional[bool] = True,
+            stream_options: Optional[Dict] = None,
             **kwargs):
         """Chat completion v1.
 
@@ -258,7 +277,11 @@ def completions_v1(
             ignore_eos (bool): indicator for ignoring eos
             skip_special_tokens (bool): Whether or not to remove special tokens
                 in the decoding. Default to be True.
-            session_id (int): Deprecated.
+            spaces_between_special_tokens (bool): Whether or not to add spaces
+                around special tokens. The behavior of Fast tokenizers is to have
+                this to False. This is setup to True in slow tokenizers.
+            stream_options: Options for streaming response. Only set this when you
+                set stream: true.
 
         Yields:
             json objects in openai formats
diff --git a/lmdeploy/serve/openai/api_server.py b/lmdeploy/serve/openai/api_server.py
index 40a3599ae..b61772f40 100644
--- a/lmdeploy/serve/openai/api_server.py
+++ b/lmdeploy/serve/openai/api_server.py
@@ -273,6 +273,8 @@ async def chat_completions_v1(request: ChatCompletionRequest, raw_request: Reque
     - n (int): How many chat completion choices to generate for each input
         message. **Only support one here**.
     - stream: whether to stream the results or not. Default to false.
+    - stream_options: Options for streaming response. Only set this when you
+        set stream: true.
     - max_tokens (int | None): output token nums. Default to None.
     - repetition_penalty (float): The parameter for repetition penalty.
         1.0 means no penalty
@@ -523,6 +525,8 @@ async def completions_v1(request: CompletionRequest, raw_request: Request = None
     - n (int): How many chat completion choices to generate for each input
         message. **Only support one here**.
     - stream: whether to stream the results or not. Default to false.
+    - stream_options: Options for streaming response. Only set this when you
+        set stream: true.
     - repetition_penalty (float): The parameter for repetition penalty.
         1.0 means no penalty
     - user (str): A unique identifier representing your end-user.