From 11a0c4142edf4987ab33ffff9412203cf5d53820 Mon Sep 17 00:00:00 2001 From: kevin Date: Wed, 18 Dec 2024 23:50:52 +0800 Subject: [PATCH 1/2] =?UTF-8?q?=E5=A2=9E=E5=8A=A0xAI=E6=A8=A1=E5=9E=8B?= =?UTF-8?q?=E6=94=AF=E6=8C=81=20=E6=8E=A8=E8=8D=90llm-models.json=E6=96=B0?= =?UTF-8?q?=E5=A2=9E=20```json=20,=20=20=20=20=20=20=20=20=20{=20=20=20=20?= =?UTF-8?q?=20=20=20=20=20=20=20=20=20"name":=20"grok-2-vision-1212",=20?= =?UTF-8?q?=20=20=20=20=20=20=20=20=20=20=20=20"model=5Fname":=20"grok-2-v?= =?UTF-8?q?ision-1212",=20=20=20=20=20=20=20=20=20=20=20=20=20"requester":?= =?UTF-8?q?=20"grok-chat-completions",=20=20=20=20=20=20=20=20=20=20=20=20?= =?UTF-8?q?=20"token=5Fmgr":=20"grok",=20=20=20=20=20=20=20=20=20=20=20=20?= =?UTF-8?q?=20"vision=5Fsupported":=20true=20=20=20=20=20=20=20=20=20}=20`?= =?UTF-8?q?``=20provider.json=20requester=E5=A2=9E=E5=8A=A0=20```json=20,?= =?UTF-8?q?=20=20=20=20=20=20=20=20=20"grok-chat-completions":=20{=20=20?= =?UTF-8?q?=20=20=20=20=20=20=20=20=20=20=20"args":=20{},=20=20=20=20=20?= =?UTF-8?q?=20=20=20=20=20=20=20=20"base-url":=20"https://api.x.ai/v1",=20?= =?UTF-8?q?=20=20=20=20=20=20=20=20=20=20=20=20"timeout":=20120=20=20=20?= =?UTF-8?q?=20=20=20=20=20=20}=20```=20keys=E5=A2=9E=E5=8A=A0:=20```json?= =?UTF-8?q?=20,=20"grok":=20[=20=20=20=20=20=20=20=20=20=20=20=20=20"xai-y?= =?UTF-8?q?our-key"=20=20=20=20=20=20=20=20=20]=20```?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit --- pkg/provider/modelmgr/modelmgr.py | 2 +- .../modelmgr/requesters/grokchatcmpl.py | 144 ++++++++++++++++++ 2 files changed, 145 insertions(+), 1 deletion(-) create mode 100644 pkg/provider/modelmgr/requesters/grokchatcmpl.py diff --git a/pkg/provider/modelmgr/modelmgr.py b/pkg/provider/modelmgr/modelmgr.py index fb6c703b..fc6a1e38 100644 --- a/pkg/provider/modelmgr/modelmgr.py +++ b/pkg/provider/modelmgr/modelmgr.py @@ -6,7 +6,7 @@ from ...core import app from . import token -from .requesters import chatcmpl, anthropicmsgs, moonshotchatcmpl, deepseekchatcmpl, ollamachat, giteeaichatcmpl +from .requesters import chatcmpl, anthropicmsgs, moonshotchatcmpl, deepseekchatcmpl, ollamachat, giteeaichatcmpl, grokchatcmpl FETCH_MODEL_LIST_URL = "https://api.qchatgpt.rockchin.top/api/v2/fetch/model_list" diff --git a/pkg/provider/modelmgr/requesters/grokchatcmpl.py b/pkg/provider/modelmgr/requesters/grokchatcmpl.py new file mode 100644 index 00000000..921d6b17 --- /dev/null +++ b/pkg/provider/modelmgr/requesters/grokchatcmpl.py @@ -0,0 +1,144 @@ +from __future__ import annotations + +import asyncio +import typing +import json +import base64 +from typing import AsyncGenerator + +import openai +import openai.types.chat.chat_completion as chat_completion +import httpx +import aiohttp +import async_lru + +from .. import entities, errors, requester +from ....core import entities as core_entities, app +from ... import entities as llm_entities +from ...tools import entities as tools_entities +from ....utils import image + + +@requester.requester_class("grok-chat-completions") +class GrokChatCompletions(requester.LLMAPIRequester): + """grok ChatCompletion API 请求器""" + + client: openai.AsyncClient + + requester_cfg: dict + + def __init__(self, ap: app.Application): + self.ap = ap + + self.requester_cfg = self.ap.provider_cfg.data['requester']['grok-chat-completions'] + + async def initialize(self): + + self.client = openai.AsyncClient( + api_key="", + base_url=self.requester_cfg['base-url'], + timeout=self.requester_cfg['timeout'], + http_client=httpx.AsyncClient( + proxies=self.ap.proxy_mgr.get_forward_proxies() + ) + ) + + async def _req( + self, + args: dict, + ) -> chat_completion.ChatCompletion: + return await self.client.chat.completions.create(**args) + + async def _make_msg( + self, + chat_completion: chat_completion.ChatCompletion, + ) -> llm_entities.Message: + chatcmpl_message = chat_completion.choices[0].message.dict() + + # 确保 role 字段存在且不为 None + if 'role' not in chatcmpl_message or chatcmpl_message['role'] is None: + chatcmpl_message['role'] = 'assistant' + + message = llm_entities.Message(**chatcmpl_message) + + return message + + async def _closure( + self, + req_messages: list[dict], + use_model: entities.LLMModelInfo, + use_funcs: list[tools_entities.LLMFunction] = None, + ) -> llm_entities.Message: + self.client.api_key = use_model.token_mgr.get_token() + + args = self.requester_cfg['args'].copy() + args["model"] = use_model.name if use_model.model_name is None else use_model.model_name + + if use_funcs: + tools = await self.ap.tool_mgr.generate_tools_for_openai(use_funcs) + + if tools: + args["tools"] = tools + + # 设置此次请求中的messages + messages = req_messages.copy() + + # 检查vision + for msg in messages: + if 'content' in msg and isinstance(msg["content"], list): + for me in msg["content"]: + if me["type"] == "image_url": + me["image_url"]['url'] = await self.get_base64_str(me["image_url"]['url']) + + args["messages"] = messages + + # 发送请求 + resp = await self._req(args) + + # 处理请求结果 + message = await self._make_msg(resp) + + return message + + async def call( + self, + model: entities.LLMModelInfo, + messages: typing.List[llm_entities.Message], + funcs: typing.List[tools_entities.LLMFunction] = None, + ) -> llm_entities.Message: + req_messages = [] # req_messages 仅用于类内，外部同步由 query.messages 进行 + for m in messages: + msg_dict = m.dict(exclude_none=True) + content = msg_dict.get("content") + if isinstance(content, list): + # 检查 content 列表中是否每个部分都是文本 + if all(isinstance(part, dict) and part.get("type") == "text" for part in content): + # 将所有文本部分合并为一个字符串 + msg_dict["content"] = "\n".join(part["text"] for part in content) + req_messages.append(msg_dict) + + try: + return await self._closure(req_messages, model, funcs) + except asyncio.TimeoutError: + raise errors.RequesterError('请求超时') + except openai.BadRequestError as e: + if 'context_length_exceeded' in e.message: + raise errors.RequesterError(f'上文过长，请重置会话: {e.message}') + else: + raise errors.RequesterError(f'请求参数错误: {e.message}') + except openai.AuthenticationError as e: + raise errors.RequesterError(f'无效的 api-key: {e.message}') + except openai.NotFoundError as e: + raise errors.RequesterError(f'请求路径错误: {e.message}') + except openai.RateLimitError as e: + raise errors.RequesterError(f'请求过于频繁或余额不足: {e.message}') + except openai.APIError as e: + raise errors.RequesterError(f'请求错误: {e.message}') + + @async_lru.alru_cache(maxsize=128) + async def get_base64_str( + self, + original_url: str, + ) -> str: + base64_image, image_format = await image.qq_image_url_to_base64(original_url) + return f"data:image/{image_format};base64,{base64_image}" From 0a68a77e28cc5925424f3fb87802646dd5bf6b0f Mon Sep 17 00:00:00 2001 From: Junyan Qin Date: Sat, 4 Jan 2025 22:24:05 +0800 Subject: [PATCH 2/2] feat: refactor --- pkg/core/migrations/m018_xai_config.py | 25 +++ pkg/core/stages/migrate.py | 2 +- pkg/provider/modelmgr/modelmgr.py | 2 +- .../modelmgr/requesters/grokchatcmpl.py | 144 ----------------- .../modelmgr/requesters/xaichatcmpl.py | 145 ++++++++++++++++++ templates/metadata/llm-models.json | 32 ++++ templates/provider.json | 8 + templates/schema/provider.json | 34 +++- 8 files changed, 240 insertions(+), 152 deletions(-) create mode 100644 pkg/core/migrations/m018_xai_config.py delete mode 100644 pkg/provider/modelmgr/requesters/grokchatcmpl.py create mode 100644 pkg/provider/modelmgr/requesters/xaichatcmpl.py diff --git a/pkg/core/migrations/m018_xai_config.py b/pkg/core/migrations/m018_xai_config.py new file mode 100644 index 00000000..bf422451 --- /dev/null +++ b/pkg/core/migrations/m018_xai_config.py @@ -0,0 +1,25 @@ +from __future__ import annotations + +from .. import migration + + +@migration.migration_class("xai-config", 18) +class XaiConfigMigration(migration.Migration): + """迁移""" + + async def need_migrate(self) -> bool: + """判断当前环境是否需要运行此迁移""" + return 'xai-chat-completions' not in self.ap.provider_cfg.data['requester'] + + async def run(self): + """执行迁移""" + self.ap.provider_cfg.data['requester']['xai-chat-completions'] = { + "base-url": "https://api.x.ai/v1", + "args": {}, + "timeout": 120 + } + self.ap.provider_cfg.data['keys']['xai'] = [ + "xai-1234567890" + ] + + await self.ap.provider_cfg.dump_config() diff --git a/pkg/core/stages/migrate.py b/pkg/core/stages/migrate.py index 3134dc3b..933cdf68 100644 --- a/pkg/core/stages/migrate.py +++ b/pkg/core/stages/migrate.py @@ -7,7 +7,7 @@ from ..migrations import m001_sensitive_word_migration, m002_openai_config_migration, m003_anthropic_requester_cfg_completion, m004_moonshot_cfg_completion from ..migrations import m005_deepseek_cfg_completion, m006_vision_config, m007_qcg_center_url, m008_ad_fixwin_config_migrate, m009_msg_truncator_cfg from ..migrations import m010_ollama_requester_config, m011_command_prefix_config, m012_runner_config, m013_http_api_config, m014_force_delay_config -from ..migrations import m015_gitee_ai_config, m016_dify_service_api, m017_dify_api_timeout_params +from ..migrations import m015_gitee_ai_config, m016_dify_service_api, m017_dify_api_timeout_params, m018_xai_config @stage.stage_class("MigrationStage") diff --git a/pkg/provider/modelmgr/modelmgr.py b/pkg/provider/modelmgr/modelmgr.py index fc6a1e38..b7ae0efe 100644 --- a/pkg/provider/modelmgr/modelmgr.py +++ b/pkg/provider/modelmgr/modelmgr.py @@ -6,7 +6,7 @@ from ...core import app from . import token -from .requesters import chatcmpl, anthropicmsgs, moonshotchatcmpl, deepseekchatcmpl, ollamachat, giteeaichatcmpl, grokchatcmpl +from .requesters import chatcmpl, anthropicmsgs, moonshotchatcmpl, deepseekchatcmpl, ollamachat, giteeaichatcmpl, xaichatcmpl FETCH_MODEL_LIST_URL = "https://api.qchatgpt.rockchin.top/api/v2/fetch/model_list" diff --git a/pkg/provider/modelmgr/requesters/grokchatcmpl.py b/pkg/provider/modelmgr/requesters/grokchatcmpl.py deleted file mode 100644 index 921d6b17..00000000 --- a/pkg/provider/modelmgr/requesters/grokchatcmpl.py +++ /dev/null @@ -1,144 +0,0 @@ -from __future__ import annotations - -import asyncio -import typing -import json -import base64 -from typing import AsyncGenerator - -import openai -import openai.types.chat.chat_completion as chat_completion -import httpx -import aiohttp -import async_lru - -from .. import entities, errors, requester -from ....core import entities as core_entities, app -from ... import entities as llm_entities -from ...tools import entities as tools_entities -from ....utils import image - - -@requester.requester_class("grok-chat-completions") -class GrokChatCompletions(requester.LLMAPIRequester): - """grok ChatCompletion API 请求器""" - - client: openai.AsyncClient - - requester_cfg: dict - - def __init__(self, ap: app.Application): - self.ap = ap - - self.requester_cfg = self.ap.provider_cfg.data['requester']['grok-chat-completions'] - - async def initialize(self): - - self.client = openai.AsyncClient( - api_key="", - base_url=self.requester_cfg['base-url'], - timeout=self.requester_cfg['timeout'], - http_client=httpx.AsyncClient( - proxies=self.ap.proxy_mgr.get_forward_proxies() - ) - ) - - async def _req( - self, - args: dict, - ) -> chat_completion.ChatCompletion: - return await self.client.chat.completions.create(**args) - - async def _make_msg( - self, - chat_completion: chat_completion.ChatCompletion, - ) -> llm_entities.Message: - chatcmpl_message = chat_completion.choices[0].message.dict() - - # 确保 role 字段存在且不为 None - if 'role' not in chatcmpl_message or chatcmpl_message['role'] is None: - chatcmpl_message['role'] = 'assistant' - - message = llm_entities.Message(**chatcmpl_message) - - return message - - async def _closure( - self, - req_messages: list[dict], - use_model: entities.LLMModelInfo, - use_funcs: list[tools_entities.LLMFunction] = None, - ) -> llm_entities.Message: - self.client.api_key = use_model.token_mgr.get_token() - - args = self.requester_cfg['args'].copy() - args["model"] = use_model.name if use_model.model_name is None else use_model.model_name - - if use_funcs: - tools = await self.ap.tool_mgr.generate_tools_for_openai(use_funcs) - - if tools: - args["tools"] = tools - - # 设置此次请求中的messages - messages = req_messages.copy() - - # 检查vision - for msg in messages: - if 'content' in msg and isinstance(msg["content"], list): - for me in msg["content"]: - if me["type"] == "image_url": - me["image_url"]['url'] = await self.get_base64_str(me["image_url"]['url']) - - args["messages"] = messages - - # 发送请求 - resp = await self._req(args) - - # 处理请求结果 - message = await self._make_msg(resp) - - return message - - async def call( - self, - model: entities.LLMModelInfo, - messages: typing.List[llm_entities.Message], - funcs: typing.List[tools_entities.LLMFunction] = None, - ) -> llm_entities.Message: - req_messages = [] # req_messages 仅用于类内，外部同步由 query.messages 进行 - for m in messages: - msg_dict = m.dict(exclude_none=True) - content = msg_dict.get("content") - if isinstance(content, list): - # 检查 content 列表中是否每个部分都是文本 - if all(isinstance(part, dict) and part.get("type") == "text" for part in content): - # 将所有文本部分合并为一个字符串 - msg_dict["content"] = "\n".join(part["text"] for part in content) - req_messages.append(msg_dict) - - try: - return await self._closure(req_messages, model, funcs) - except asyncio.TimeoutError: - raise errors.RequesterError('请求超时') - except openai.BadRequestError as e: - if 'context_length_exceeded' in e.message: - raise errors.RequesterError(f'上文过长，请重置会话: {e.message}') - else: - raise errors.RequesterError(f'请求参数错误: {e.message}') - except openai.AuthenticationError as e: - raise errors.RequesterError(f'无效的 api-key: {e.message}') - except openai.NotFoundError as e: - raise errors.RequesterError(f'请求路径错误: {e.message}') - except openai.RateLimitError as e: - raise errors.RequesterError(f'请求过于频繁或余额不足: {e.message}') - except openai.APIError as e: - raise errors.RequesterError(f'请求错误: {e.message}') - - @async_lru.alru_cache(maxsize=128) - async def get_base64_str( - self, - original_url: str, - ) -> str: - base64_image, image_format = await image.qq_image_url_to_base64(original_url) - return f"data:image/{image_format};base64,{base64_image}" diff --git a/pkg/provider/modelmgr/requesters/xaichatcmpl.py b/pkg/provider/modelmgr/requesters/xaichatcmpl.py new file mode 100644 index 00000000..89578179 --- /dev/null +++ b/pkg/provider/modelmgr/requesters/xaichatcmpl.py @@ -0,0 +1,145 @@ +from __future__ import annotations + +import asyncio +import typing +import json +import base64 +from typing import AsyncGenerator + +import openai +import openai.types.chat.chat_completion as chat_completion +import httpx +import aiohttp +import async_lru + +from . import chatcmpl +from .. import entities, errors, requester +from ....core import entities as core_entities, app +from ... import entities as llm_entities +from ...tools import entities as tools_entities +from ....utils import image + + +@requester.requester_class("xai-chat-completions") +class XaiChatCompletions(chatcmpl.OpenAIChatCompletions): + """xAI ChatCompletion API 请求器""" + + client: openai.AsyncClient + + requester_cfg: dict + + def __init__(self, ap: app.Application): + self.ap = ap + + self.requester_cfg = self.ap.provider_cfg.data['requester']['xai-chat-completions'] + + # async def initialize(self): + + # self.client = openai.AsyncClient( + # api_key="", + # base_url=self.requester_cfg['base-url'], + # timeout=self.requester_cfg['timeout'], + # http_client=httpx.AsyncClient( + # proxies=self.ap.proxy_mgr.get_forward_proxies() + # ) + # ) + + # async def _req( + # self, + # args: dict, + # ) -> chat_completion.ChatCompletion: + # return await self.client.chat.completions.create(**args) + + # async def _make_msg( + # self, + # chat_completion: chat_completion.ChatCompletion, + # ) -> llm_entities.Message: + # chatcmpl_message = chat_completion.choices[0].message.dict() + + # # 确保 role 字段存在且不为 None + # if 'role' not in chatcmpl_message or chatcmpl_message['role'] is None: + # chatcmpl_message['role'] = 'assistant' + + # message = llm_entities.Message(**chatcmpl_message) + + # return message + + # async def _closure( + # self, + # req_messages: list[dict], + # use_model: entities.LLMModelInfo, + # use_funcs: list[tools_entities.LLMFunction] = None, + # ) -> llm_entities.Message: + # self.client.api_key = use_model.token_mgr.get_token() + + # args = self.requester_cfg['args'].copy() + # args["model"] = use_model.name if use_model.model_name is None else use_model.model_name + + # if use_funcs: + # tools = await self.ap.tool_mgr.generate_tools_for_openai(use_funcs) + + # if tools: + # args["tools"] = tools + + # # 设置此次请求中的messages + # messages = req_messages.copy() + + # # 检查vision + # for msg in messages: + # if 'content' in msg and isinstance(msg["content"], list): + # for me in msg["content"]: + # if me["type"] == "image_url": + # me["image_url"]['url'] = await self.get_base64_str(me["image_url"]['url']) + + # args["messages"] = messages + + # # 发送请求 + # resp = await self._req(args) + + # # 处理请求结果 + # message = await self._make_msg(resp) + + # return message + + # async def call( + # self, + # model: entities.LLMModelInfo, + # messages: typing.List[llm_entities.Message], + # funcs: typing.List[tools_entities.LLMFunction] = None, + # ) -> llm_entities.Message: + # req_messages = [] # req_messages 仅用于类内，外部同步由 query.messages 进行 + # for m in messages: + # msg_dict = m.dict(exclude_none=True) + # content = msg_dict.get("content") + # if isinstance(content, list): + # # 检查 content 列表中是否每个部分都是文本 + # if all(isinstance(part, dict) and part.get("type") == "text" for part in content): + # # 将所有文本部分合并为一个字符串 + # msg_dict["content"] = "\n".join(part["text"] for part in content) + # req_messages.append(msg_dict) + + # try: + # return await self._closure(req_messages, model, funcs) + # except asyncio.TimeoutError: + # raise errors.RequesterError('请求超时') + # except openai.BadRequestError as e: + # if 'context_length_exceeded' in e.message: + # raise errors.RequesterError(f'上文过长，请重置会话: {e.message}') + # else: + # raise errors.RequesterError(f'请求参数错误: {e.message}') + # except openai.AuthenticationError as e: + # raise errors.RequesterError(f'无效的 api-key: {e.message}') + # except openai.NotFoundError as e: + # raise errors.RequesterError(f'请求路径错误: {e.message}') + # except openai.RateLimitError as e: + # raise errors.RequesterError(f'请求过于频繁或余额不足: {e.message}') + # except openai.APIError as e: + # raise errors.RequesterError(f'请求错误: {e.message}') + + # @async_lru.alru_cache(maxsize=128) + # async def get_base64_str( + # self, + # original_url: str, + # ) -> str: + # base64_image, image_format = await image.qq_image_url_to_base64(original_url) + # return f"data:image/{image_format};base64,{base64_image}" diff --git a/templates/metadata/llm-models.json b/templates/metadata/llm-models.json index 7810aacc..6e6bc1f5 100644 --- a/templates/metadata/llm-models.json +++ b/templates/metadata/llm-models.json @@ -115,6 +115,38 @@ "name": "deepseek-coder", "requester": "deepseek-chat-completions", "token_mgr": "deepseek" + }, + { + "name": "grok-2-latest", + "requester": "xai-chat-completions", + "token_mgr": "xai" + }, + { + "name": "grok-2", + "requester": "xai-chat-completions", + "token_mgr": "xai" + }, + { + "name": "grok-2-vision-1212", + "requester": "xai-chat-completions", + "token_mgr": "xai", + "vision_supported": true + }, + { + "name": "grok-2-1212", + "requester": "xai-chat-completions", + "token_mgr": "xai" + }, + { + "name": "grok-vision-beta", + "requester": "xai-chat-completions", + "token_mgr": "xai", + "vision_supported": true + }, + { + "name": "grok-beta", + "requester": "xai-chat-completions", + "token_mgr": "xai" } ] } \ No newline at end of file diff --git a/templates/provider.json b/templates/provider.json index 30656f8c..f1413ef7 100644 --- a/templates/provider.json +++ b/templates/provider.json @@ -16,6 +16,9 @@ ], "gitee-ai": [ "XXXXX" + ], + "xai": [ + "xai-1234567890" ] }, "requester": { @@ -50,6 +53,11 @@ "base-url": "https://ai.gitee.com/v1", "args": {}, "timeout": 120 + }, + "xai-chat-completions": { + "base-url": "https://api.x.ai/v1", + "args": {}, + "timeout": 120 } }, "model": "gpt-4o", diff --git a/templates/schema/provider.json b/templates/schema/provider.json index d4e453a4..1b819644 100644 --- a/templates/schema/provider.json +++ b/templates/schema/provider.json @@ -22,7 +22,6 @@ "openai": { "type": "array", "title": "OpenAI API 密钥", - "description": "OpenAI API 密钥", "items": { "type": "string" }, @@ -31,7 +30,6 @@ "anthropic": { "type": "array", "title": "Anthropic API 密钥", - "description": "Anthropic API 密钥", "items": { "type": "string" }, @@ -40,7 +38,6 @@ "moonshot": { "type": "array", "title": "Moonshot API 密钥", - "description": "Moonshot API 密钥", "items": { "type": "string" }, @@ -49,7 +46,6 @@ "deepseek": { "type": "array", "title": "DeepSeek API 密钥", - "description": "DeepSeek API 密钥", "items": { "type": "string" }, @@ -57,8 +53,15 @@ }, "gitee": { "type": "array", - "title": "Gitee API 密钥", - "description": "Gitee API 密钥", + "title": "Gitee AI API 密钥", + "items": { + "type": "string" + }, + "default": [] + }, + "xai": { + "type": "array", + "title": "xAI API 密钥", "items": { "type": "string" }, @@ -188,6 +191,25 @@ "default": 120 } } + }, + "xai-chat-completions": { + "type": "object", + "title": "xAI API 请求配置", + "description": "仅可编辑 URL 和 超时时间，额外请求参数不支持可视化编辑，请到编辑器编辑", + "properties": { + "base-url": { + "type": "string", + "title": "API URL" + }, + "args": { + "type": "object" + }, + "timeout": { + "type": "number", + "title": "API 请求超时时间", + "default": 120 + } + } } } },