From 4a27c9259dccb1c0ad7e7a288131f9698ddbb37e Mon Sep 17 00:00:00 2001 From: yongtenglei Date: Mon, 1 Dec 2025 17:45:21 +0800 Subject: [PATCH] backend --- conf/llm_factories.json | 33 ++--------------- rag/llm/__init__.py | 3 ++ rag/llm/chat_model.py | 82 +---------------------------------------- 3 files changed, 8 insertions(+), 110 deletions(-) diff --git a/conf/llm_factories.json b/conf/llm_factories.json index 3c84bd03d..261725ea0 100644 --- a/conf/llm_factories.json +++ b/conf/llm_factories.json @@ -1232,39 +1232,14 @@ { "name": "MiniMax", "logo": "", - "tags": "LLM,TEXT EMBEDDING", + "tags": "LLM", "status": "1", "rank": "810", "llm": [ { - "llm_name": "abab6.5-chat", - "tags": "LLM,CHAT,8k", - "max_tokens": 8192, - "model_type": "chat" - }, - { - "llm_name": "abab6.5s-chat", - "tags": "LLM,CHAT,245k", - "max_tokens": 245760, - "model_type": "chat", - "is_tools": true - }, - { - "llm_name": "abab6.5t-chat", - "tags": "LLM,CHAT,8k", - "max_tokens": 8192, - "model_type": "chat" - }, - { - "llm_name": "abab6.5g-chat", - "tags": "LLM,CHAT,8k", - "max_tokens": 8192, - "model_type": "chat" - }, - { - "llm_name": "abab5.5s-chat", - "tags": "LLM,CHAT,8k", - "max_tokens": 8192, + "llm_name": "MiniMax-M2", + "tags": "LLM,CHAT,200k", + "max_tokens": 200000, "model_type": "chat" } ] diff --git a/rag/llm/__init__.py b/rag/llm/__init__.py index 1913646a2..3ff5311fc 100644 --- a/rag/llm/__init__.py +++ b/rag/llm/__init__.py @@ -51,6 +51,7 @@ class SupportedLiteLLMProvider(StrEnum): AI_302 = "302.AI" JiekouAI = "Jiekou.AI" ZHIPU_AI = "ZHIPU-AI" + MiniMax = "MiniMax" FACTORY_DEFAULT_BASE_URL = { @@ -73,6 +74,7 @@ FACTORY_DEFAULT_BASE_URL = { SupportedLiteLLMProvider.Anthropic: "https://api.anthropic.com/", SupportedLiteLLMProvider.JiekouAI: "https://api.jiekou.ai/openai", SupportedLiteLLMProvider.ZHIPU_AI: "https://open.bigmodel.cn/api/paas/v4", + SupportedLiteLLMProvider.MiniMax: "https://api.minimaxi.com/v1", } @@ -105,6 +107,7 @@ LITELLM_PROVIDER_PREFIX = { SupportedLiteLLMProvider.AI_302: "openai/", SupportedLiteLLMProvider.JiekouAI: "openai/", SupportedLiteLLMProvider.ZHIPU_AI: "openai/", + SupportedLiteLLMProvider.MiniMax: "openai/", } ChatModel = globals().get("ChatModel", {}) diff --git a/rag/llm/chat_model.py b/rag/llm/chat_model.py index 1f38292ba..1b7140a2b 100644 --- a/rag/llm/chat_model.py +++ b/rag/llm/chat_model.py @@ -28,7 +28,6 @@ from urllib.parse import urljoin import json_repair import litellm import openai -import requests from openai import AsyncOpenAI, OpenAI from openai.lib.azure import AzureOpenAI from strenum import StrEnum @@ -1015,86 +1014,6 @@ class VolcEngineChat(Base): super().__init__(ark_api_key, model_name, base_url, **kwargs) -class MiniMaxChat(Base): - _FACTORY_NAME = "MiniMax" - - def __init__(self, key, model_name, base_url="https://api.minimax.chat/v1/text/chatcompletion_v2", **kwargs): - super().__init__(key, model_name, base_url=base_url, **kwargs) - - if not base_url: - base_url = "https://api.minimax.chat/v1/text/chatcompletion_v2" - self.base_url = base_url - self.model_name = model_name - self.api_key = key - - def _clean_conf(self, gen_conf): - for k in list(gen_conf.keys()): - if k not in ["temperature", "top_p", "max_tokens"]: - del gen_conf[k] - return gen_conf - - def _chat(self, history, gen_conf): - headers = { - "Authorization": f"Bearer {self.api_key}", - "Content-Type": "application/json", - } - payload = json.dumps({"model": self.model_name, "messages": history, **gen_conf}) - response = requests.request("POST", url=self.base_url, headers=headers, data=payload) - response = response.json() - ans = response["choices"][0]["message"]["content"].strip() - if response["choices"][0]["finish_reason"] == "length": - if is_chinese(ans): - ans += LENGTH_NOTIFICATION_CN - else: - ans += LENGTH_NOTIFICATION_EN - return ans, total_token_count_from_response(response) - - def chat_streamly(self, system, history, gen_conf): - if system and history and history[0].get("role") != "system": - history.insert(0, {"role": "system", "content": system}) - for k in list(gen_conf.keys()): - if k not in ["temperature", "top_p", "max_tokens"]: - del gen_conf[k] - ans = "" - total_tokens = 0 - try: - headers = { - "Authorization": f"Bearer {self.api_key}", - "Content-Type": "application/json", - } - payload = json.dumps( - { - "model": self.model_name, - "messages": history, - "stream": True, - **gen_conf, - } - ) - response = requests.request( - "POST", - url=self.base_url, - headers=headers, - data=payload, - ) - for resp in response.text.split("\n\n")[:-1]: - resp = json.loads(resp[6:]) - text = "" - if "choices" in resp and "delta" in resp["choices"][0]: - text = resp["choices"][0]["delta"]["content"] - ans = text - tol = total_token_count_from_response(resp) - if not tol: - total_tokens += num_tokens_from_string(text) - else: - total_tokens = tol - yield ans - - except Exception as e: - yield ans + "\n**ERROR**: " + str(e) - - yield total_tokens - - class MistralChat(Base): _FACTORY_NAME = "Mistral" @@ -1642,6 +1561,7 @@ class LiteLLMBase(ABC): "302.AI", "Jiekou.AI", "ZHIPU-AI", + "MiniMax", ] def __init__(self, key, model_name, base_url=None, **kwargs):