From 0b5d1d3d18736899d5665af846a2eb88873bc8f0 Mon Sep 17 00:00:00 2001 From: Bagus Indrayana Date: Thu, 17 Aug 2023 23:36:33 +0800 Subject: add provider --- g4f/Provider/Wuguokai.py | 65 ++++++++++++++++++++++++++++++++++++++++++++++++ g4f/Provider/__init__.py | 4 ++- 2 files changed, 68 insertions(+), 1 deletion(-) create mode 100644 g4f/Provider/Wuguokai.py (limited to 'g4f') diff --git a/g4f/Provider/Wuguokai.py b/g4f/Provider/Wuguokai.py new file mode 100644 index 00000000..906283ad --- /dev/null +++ b/g4f/Provider/Wuguokai.py @@ -0,0 +1,65 @@ +import random, requests, json +from ..typing import Any, CreateResult +from .base_provider import BaseProvider + + +class Wuguokai(BaseProvider): + url = 'https://chat.wuguokai.xyz' + supports_gpt_35_turbo = True + supports_stream = False + needs_auth = False + working = True + + @staticmethod + def create_completion( + model: str, + messages: list[dict[str, str]], + stream: bool, + **kwargs: Any, + ) -> CreateResult: + base = '' + for message in messages: + base += '%s: %s\n' % (message['role'], message['content']) + base += 'assistant:' + + headers = { + 'authority': 'ai-api.wuguokai.xyz', + 'accept': 'application/json, text/plain, */*', + 'accept-language': 'id-ID,id;q=0.9,en-US;q=0.8,en;q=0.7', + 'content-type': 'application/json', + 'origin': 'https://chat.wuguokai.xyz', + 'referer': 'https://chat.wuguokai.xyz/', + 'sec-ch-ua': '"Not.A/Brand";v="8", "Chromium";v="114", "Google Chrome";v="114"', + 'sec-ch-ua-mobile': '?0', + 'sec-ch-ua-platform': '"Windows"', + 'sec-fetch-dest': 'empty', + 'sec-fetch-mode': 'cors', + 'sec-fetch-site': 'same-site', + 'user-agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/114.0.0.0 Safari/537.36' + } + data ={ + "prompt": base, + "options": {}, + "userId": f"#/chat/{random.randint(1,99999999)}", + "usingContext": True + } + response = requests.post("https://ai-api20.wuguokai.xyz/api/chat-process", headers=headers, data=json.dumps(data),proxies=kwargs['proxy'] if 'proxy' in kwargs else {}) + _split = response.text.split("> 若回答失败请重试或多刷新几次界面后重试") + if response.status_code == 200: + if len(_split) > 1: + yield _split[1].strip() + else: + yield _split[0].strip() + else: + raise Exception(f"Error: {response.status_code} {response.reason}") + + @classmethod + @property + def params(cls): + params = [ + ("model", "str"), + ("messages", "list[dict[str, str]]"), + ("stream", "bool") + ] + param = ", ".join([": ".join(p) for p in params]) + return f"g4f.provider.{cls.__name__} supports: ({param})" \ No newline at end of file diff --git a/g4f/Provider/__init__.py b/g4f/Provider/__init__.py index 81d1ad64..432ae672 100644 --- a/g4f/Provider/__init__.py +++ b/g4f/Provider/__init__.py @@ -25,6 +25,7 @@ from .You import You from .Yqcloud import Yqcloud from .Equing import Equing from .FastGpt import FastGpt +from .Wuguokai import Wuguokai __all__ = [ "BaseProvider", @@ -53,5 +54,6 @@ __all__ = [ "You", "Yqcloud", "Equing", - "FastGpt" + "FastGpt", + "Wuguokai" ] -- cgit v1.2.3 From ce45d36936b4d33ebc41c6687095ca511a175a35 Mon Sep 17 00:00:00 2001 From: Heiner Lohaus Date: Mon, 21 Aug 2023 22:39:57 +0200 Subject: Fix bing provider --- g4f/Provider/Bing.py | 464 ++++++++++++++++++++++----------------------------- 1 file changed, 203 insertions(+), 261 deletions(-) (limited to 'g4f') diff --git a/g4f/Provider/Bing.py b/g4f/Provider/Bing.py index 48b5477d..364e9c1d 100644 --- a/g4f/Provider/Bing.py +++ b/g4f/Provider/Bing.py @@ -2,17 +2,15 @@ import asyncio import json import os import random -import ssl -import uuid import aiohttp -import certifi -import requests +import asyncio +import browser_cookie3 +from aiohttp import ClientSession -from ..typing import Any, AsyncGenerator, CreateResult, Tuple, Union +from ..typing import Any, AsyncGenerator, CreateResult, Union from .base_provider import BaseProvider - class Bing(BaseProvider): url = "https://bing.com/chat" supports_gpt_4 = True @@ -22,20 +20,27 @@ class Bing(BaseProvider): model: str, messages: list[dict[str, str]], stream: bool, - **kwargs: Any, + **kwargs: Any ) -> CreateResult: - if len(messages) < 2: - prompt = messages[0]["content"] - context = False + yield from run(create(messages, **kwargs)) + +def create( + messages: list[dict[str, str]], + cookies: dict = {} + ): + if len(messages) < 2: + prompt = messages[0]["content"] + context = None - else: - prompt = messages[-1]["content"] - context = convert(messages[:-1]) + else: + prompt = messages[-1]["content"] + context = convert(messages[:-1]) - response = run(stream_generate(prompt, jailbreak, context)) - for token in response: - yield token + if not cookies: + for cookie in browser_cookie3.load(domain_name='.bing.com'): + cookies[cookie.name] = cookie.value + return stream_generate(prompt, context, cookies) def convert(messages: list[dict[str, str]]): context = "" @@ -45,250 +50,43 @@ def convert(messages: list[dict[str, str]]): return context - -jailbreak = { - "optionsSets": [ - "saharasugg", - "enablenewsfc", - "clgalileo", - "gencontentv3", - "nlu_direct_response_filter", - "deepleo", - "disable_emoji_spoken_text", - "responsible_ai_policy_235", - "enablemm", - "h3precise" - # "harmonyv3", - "dtappid", - "cricinfo", - "cricinfov2", - "dv3sugg", - "nojbfedge", - ] -} - - -ssl_context = ssl.create_default_context() -ssl_context.load_verify_locations(certifi.where()) - - -def _format(msg: dict[str, Any]) -> str: - return json.dumps(msg, ensure_ascii=False) + Defaults.delimiter - - -async def stream_generate( - prompt: str, - mode: dict[str, list[str]] = jailbreak, - context: Union[bool, str] = False, -): - timeout = aiohttp.ClientTimeout(total=900) - session = aiohttp.ClientSession(timeout=timeout) - - conversationId, clientId, conversationSignature = await create_conversation() - - wss = await session.ws_connect( - "wss://sydney.bing.com/sydney/ChatHub", - ssl=ssl_context, - autoping=False, - headers={ - "accept": "application/json", - "accept-language": "en-US,en;q=0.9", - "content-type": "application/json", - "sec-ch-ua": '"Not_A Brand";v="99", "Microsoft Edge";v="110", "Chromium";v="110"', - "sec-ch-ua-arch": '"x86"', - "sec-ch-ua-bitness": '"64"', - "sec-ch-ua-full-version": '"109.0.1518.78"', - "sec-ch-ua-full-version-list": '"Chromium";v="110.0.5481.192", "Not A(Brand";v="24.0.0.0", "Microsoft Edge";v="110.0.1587.69"', - "sec-ch-ua-mobile": "?0", - "sec-ch-ua-model": "", - "sec-ch-ua-platform": '"Windows"', - "sec-ch-ua-platform-version": '"15.0.0"', - "sec-fetch-dest": "empty", - "sec-fetch-mode": "cors", - "sec-fetch-site": "same-origin", - "x-ms-client-request-id": str(uuid.uuid4()), - "x-ms-useragent": "azsdk-js-api-client-factory/1.0.0-beta.1 core-rest-pipeline/1.10.0 OS/Win32", - "Referer": "https://www.bing.com/search?q=Bing+AI&showconv=1&FORM=hpcodx", - "Referrer-Policy": "origin-when-cross-origin", - "x-forwarded-for": Defaults.ip_address, - }, - ) - - await wss.send_str(_format({"protocol": "json", "version": 1})) - await wss.receive(timeout=900) - - argument: dict[str, Any] = { - **mode, +class Conversation(): + def __init__(self, conversationId: str, clientId: str, conversationSignature: str) -> None: + self.conversationId = conversationId + self.clientId = clientId + self.conversationSignature = conversationSignature + +async def create_conversation(session: ClientSession) -> Conversation: + url = 'https://www.bing.com/turing/conversation/create' + async with await session.get(url) as response: + response = await response.json() + conversationId = response.get('conversationId') + clientId = response.get('clientId') + conversationSignature = response.get('conversationSignature') + + if not conversationId or not clientId or not conversationSignature: + raise Exception('Failed to create conversation.') + + return Conversation(conversationId, clientId, conversationSignature) + +async def list_conversations(session: ClientSession) -> list: + url = "https://www.bing.com/turing/conversation/chats" + async with session.get(url) as response: + response = await response.json() + return response["chats"] + +async def delete_conversation(session: ClientSession, conversation: Conversation) -> list: + url = "https://sydney.bing.com/sydney/DeleteSingleConversation" + json = { + "conversationId": conversation.conversationId, + "conversationSignature": conversation.conversationSignature, + "participant": {"id": conversation.clientId}, "source": "cib", - "allowedMessageTypes": Defaults.allowedMessageTypes, - "sliceIds": Defaults.sliceIds, - "traceId": os.urandom(16).hex(), - "isStartOfSession": True, - "message": Defaults.location - | { - "author": "user", - "inputMethod": "Keyboard", - "text": prompt, - "messageType": "Chat", - }, - "conversationSignature": conversationSignature, - "participant": {"id": clientId}, - "conversationId": conversationId, + "optionsSets": ["autosave"] } - - if context: - argument["previousMessages"] = [ - { - "author": "user", - "description": context, - "contextType": "WebPage", - "messageType": "Context", - "messageId": "discover-web--page-ping-mriduna-----", - } - ] - - struct: dict[str, list[dict[str, Any]] | str | int] = { - "arguments": [argument], - "invocationId": "0", - "target": "chat", - "type": 4, - } - - await wss.send_str(_format(struct)) - - final = False - draw = False - resp_txt = "" - result_text = "" - resp_txt_no_link = "" - cache_text = "" - - while not final: - msg = await wss.receive(timeout=900) - objects = msg.data.split(Defaults.delimiter) # type: ignore - - for obj in objects: # type: ignore - if obj is None or not obj: - continue - - response = json.loads(obj) # type: ignore - if response.get("type") == 1 and response["arguments"][0].get( - "messages", - ): - if not draw: - if ( - response["arguments"][0]["messages"][0]["contentOrigin"] - != "Apology" - ) and not draw: - resp_txt = result_text + response["arguments"][0]["messages"][ - 0 - ]["adaptiveCards"][0]["body"][0].get("text", "") - resp_txt_no_link = result_text + response["arguments"][0][ - "messages" - ][0].get("text", "") - - if response["arguments"][0]["messages"][0].get( - "messageType", - ): - resp_txt = ( - resp_txt - + response["arguments"][0]["messages"][0][ - "adaptiveCards" - ][0]["body"][0]["inlines"][0].get("text") - + "\n" - ) - result_text = ( - result_text - + response["arguments"][0]["messages"][0][ - "adaptiveCards" - ][0]["body"][0]["inlines"][0].get("text") - + "\n" - ) - - if cache_text.endswith(" "): - final = True - if wss and not wss.closed: - await wss.close() - if session and not session.closed: - await session.close() - - yield (resp_txt.replace(cache_text, "")) - cache_text = resp_txt - - elif response.get("type") == 2: - if response["item"]["result"].get("error"): - if wss and not wss.closed: - await wss.close() - if session and not session.closed: - await session.close() - - raise Exception( - f"{response['item']['result']['value']}: {response['item']['result']['message']}" - ) - - if draw: - cache = response["item"]["messages"][1]["adaptiveCards"][0]["body"][ - 0 - ]["text"] - response["item"]["messages"][1]["adaptiveCards"][0]["body"][0][ - "text" - ] = (cache + resp_txt) - - if ( - response["item"]["messages"][-1]["contentOrigin"] == "Apology" - and resp_txt - ): - response["item"]["messages"][-1]["text"] = resp_txt_no_link - response["item"]["messages"][-1]["adaptiveCards"][0]["body"][0][ - "text" - ] = resp_txt - - # print('Preserved the message from being deleted', file=sys.stderr) - - final = True - if wss and not wss.closed: - await wss.close() - if session and not session.closed: - await session.close() - - -async def create_conversation() -> Tuple[str, str, str]: - create = requests.get( - "https://www.bing.com/turing/conversation/create", - headers={ - "authority": "edgeservices.bing.com", - "accept": "text/html,application/xhtml+xml,application/xml;q=0.9,image/webp,image/apng,*/*;q=0.8,application/signed-exchange;v=b3;q=0.7", - "accept-language": "en-US,en;q=0.9", - "cache-control": "max-age=0", - "sec-ch-ua": '"Chromium";v="110", "Not A(Brand";v="24", "Microsoft Edge";v="110"', - "sec-ch-ua-arch": '"x86"', - "sec-ch-ua-bitness": '"64"', - "sec-ch-ua-full-version": '"110.0.1587.69"', - "sec-ch-ua-full-version-list": '"Chromium";v="110.0.5481.192", "Not A(Brand";v="24.0.0.0", "Microsoft Edge";v="110.0.1587.69"', - "sec-ch-ua-mobile": "?0", - "sec-ch-ua-model": '""', - "sec-ch-ua-platform": '"Windows"', - "sec-ch-ua-platform-version": '"15.0.0"', - "sec-fetch-dest": "document", - "sec-fetch-mode": "navigate", - "sec-fetch-site": "none", - "sec-fetch-user": "?1", - "upgrade-insecure-requests": "1", - "user-agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/110.0.0.0 Safari/537.36 Edg/110.0.1587.69", - "x-edge-shopping-flag": "1", - "x-forwarded-for": Defaults.ip_address, - }, - ) - - conversationId = create.json().get("conversationId") - clientId = create.json().get("clientId") - conversationSignature = create.json().get("conversationSignature") - - if not conversationId or not clientId or not conversationSignature: - raise Exception("Failed to create conversation.") - - return conversationId, clientId, conversationSignature - + async with session.post(url, json=json) as response: + response = await response.json() + return response["result"]["value"] == "Success" class Defaults: delimiter = "\x1e" @@ -309,9 +107,6 @@ class Defaults: ] sliceIds = [ - # "222dtappid", - # "225cricinfo", - # "224locals0" "winmuid3tf", "osbsdusgreccf", "ttstmout", @@ -349,6 +144,153 @@ class Defaults: ], } + headers = { + 'accept': '*/*', + 'accept-language': 'en-US,en;q=0.9', + 'cache-control': 'max-age=0', + 'sec-ch-ua': '"Chromium";v="110", "Not A(Brand";v="24", "Microsoft Edge";v="110"', + 'sec-ch-ua-arch': '"x86"', + 'sec-ch-ua-bitness': '"64"', + 'sec-ch-ua-full-version': '"110.0.1587.69"', + 'sec-ch-ua-full-version-list': '"Chromium";v="110.0.5481.192", "Not A(Brand";v="24.0.0.0", "Microsoft Edge";v="110.0.1587.69"', + 'sec-ch-ua-mobile': '?0', + 'sec-ch-ua-model': '""', + 'sec-ch-ua-platform': '"Windows"', + 'sec-ch-ua-platform-version': '"15.0.0"', + 'sec-fetch-dest': 'document', + 'sec-fetch-mode': 'navigate', + 'sec-fetch-site': 'none', + 'sec-fetch-user': '?1', + 'upgrade-insecure-requests': '1', + 'user-agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/110.0.0.0 Safari/537.36 Edg/110.0.1587.69', + 'x-edge-shopping-flag': '1', + 'x-forwarded-for': ip_address, + } + + optionsSets = { + "optionsSets": [ + 'saharasugg', + 'enablenewsfc', + 'clgalileo', + 'gencontentv3', + "nlu_direct_response_filter", + "deepleo", + "disable_emoji_spoken_text", + "responsible_ai_policy_235", + "enablemm", + "h3precise" + "dtappid", + "cricinfo", + "cricinfov2", + "dv3sugg", + "nojbfedge" + ] + } + +def format_message(msg: dict) -> str: + return json.dumps(msg, ensure_ascii=False) + Defaults.delimiter + +def create_message(conversation: Conversation, prompt: str, context: str=None) -> str: + struct = { + 'arguments': [ + { + **Defaults.optionsSets, + 'source': 'cib', + 'allowedMessageTypes': Defaults.allowedMessageTypes, + 'sliceIds': Defaults.sliceIds, + 'traceId': os.urandom(16).hex(), + 'isStartOfSession': True, + 'message': Defaults.location | { + 'author': 'user', + 'inputMethod': 'Keyboard', + 'text': prompt, + 'messageType': 'Chat' + }, + 'conversationSignature': conversation.conversationSignature, + 'participant': { + 'id': conversation.clientId + }, + 'conversationId': conversation.conversationId + } + ], + 'invocationId': '0', + 'target': 'chat', + 'type': 4 + } + + if context: + struct['arguments'][0]['previousMessages'] = [{ + "author": "user", + "description": context, + "contextType": "WebPage", + "messageType": "Context", + "messageId": "discover-web--page-ping-mriduna-----" + }] + return format_message(struct) + +async def stream_generate( + prompt: str, + context: str=None, + cookies: dict=None + ): + async with ClientSession( + timeout=aiohttp.ClientTimeout(total=900), + cookies=cookies, + headers=Defaults.headers, + ) as session: + conversation = await create_conversation(session) + try: + async with session.ws_connect( + 'wss://sydney.bing.com/sydney/ChatHub', + autoping=False, + ) as wss: + + await wss.send_str(format_message({'protocol': 'json', 'version': 1})) + msg = await wss.receive(timeout=900) + + await wss.send_str(create_message(conversation, prompt, context)) + + response_txt = '' + result_text = '' + returned_text = '' + final = False + + while not final: + msg = await wss.receive(timeout=900) + objects = msg.data.split(Defaults.delimiter) + for obj in objects: + if obj is None or not obj: + continue + + response = json.loads(obj) + if response.get('type') == 1 and response['arguments'][0].get('messages'): + message = response['arguments'][0]['messages'][0] + if (message['contentOrigin'] != 'Apology'): + response_txt = result_text + \ + message['adaptiveCards'][0]['body'][0].get('text', '') + + if message.get('messageType'): + inline_txt = message['adaptiveCards'][0]['body'][0]['inlines'][0].get('text') + response_txt += inline_txt + '\n' + result_text += inline_txt + '\n' + + if returned_text.endswith(' '): + final = True + break + + if response_txt.startswith(returned_text): + new = response_txt[len(returned_text):] + if new != "\n": + yield new + returned_text = response_txt + elif response.get('type') == 2: + result = response['item']['result'] + if result.get('error'): + raise Exception(f"{result['value']}: {result['message']}") + final = True + break + finally: + await delete_conversation(session, conversation) def run(generator: AsyncGenerator[Union[Any, str], Any]): loop = asyncio.get_event_loop() -- cgit v1.2.3 From d1b6921b5fa3d9dbd7f1cce6e4f396744a7a2274 Mon Sep 17 00:00:00 2001 From: Heiner Lohaus Date: Wed, 23 Aug 2023 02:16:35 +0200 Subject: Add create_async method --- g4f/Provider/Bing.py | 58 +++++++++++++++++++++++++++++++++++----------------- 1 file changed, 39 insertions(+), 19 deletions(-) (limited to 'g4f') diff --git a/g4f/Provider/Bing.py b/g4f/Provider/Bing.py index 364e9c1d..b9b9e9a4 100644 --- a/g4f/Provider/Bing.py +++ b/g4f/Provider/Bing.py @@ -13,34 +13,54 @@ from .base_provider import BaseProvider class Bing(BaseProvider): url = "https://bing.com/chat" + working = True supports_gpt_4 = True - @staticmethod + @classmethod def create_completion( + cls, model: str, messages: list[dict[str, str]], stream: bool, **kwargs: Any ) -> CreateResult: - yield from run(create(messages, **kwargs)) - -def create( + if stream: + yield from run(cls.create_async_generator(model, messages, **kwargs)) + else: + yield asyncio.run(cls.create_async(model, messages, **kwargs)) + + @classmethod + async def create_async( + cls, + model: str, messages: list[dict[str, str]], - cookies: dict = {} - ): - if len(messages) < 2: - prompt = messages[0]["content"] - context = None - - else: - prompt = messages[-1]["content"] - context = convert(messages[:-1]) - - if not cookies: - for cookie in browser_cookie3.load(domain_name='.bing.com'): - cookies[cookie.name] = cookie.value - - return stream_generate(prompt, context, cookies) + **kwargs: Any, + ) -> str: + result = [] + async for chunk in cls.create_async_generator(model, messages, **kwargs): + result.append(chunk) + if result: + return "".join(result) + + @staticmethod + def create_async_generator( + model: str, + messages: list[dict[str, str]], + cookies: dict = {} + ) -> AsyncGenerator: + if len(messages) < 2: + prompt = messages[0]["content"] + context = None + + else: + prompt = messages[-1]["content"] + context = convert(messages[:-1]) + + if not cookies: + for cookie in browser_cookie3.load(domain_name='.bing.com'): + cookies[cookie.name] = cookie.value + + return stream_generate(prompt, context, cookies) def convert(messages: list[dict[str, str]]): context = "" -- cgit v1.2.3 From 7222d9b7cea4db173258c1189f5069b544606eeb Mon Sep 17 00:00:00 2001 From: zengrr <47846202+zeng-rr@users.noreply.github.com> Date: Wed, 23 Aug 2023 15:26:00 +0800 Subject: "You" unicode decode and fix prompt just is last message --- g4f/Provider/You.py | 13 ++++++++----- 1 file changed, 8 insertions(+), 5 deletions(-) (limited to 'g4f') diff --git a/g4f/Provider/You.py b/g4f/Provider/You.py index 0d8114a8..a8de7dec 100644 --- a/g4f/Provider/You.py +++ b/g4f/Provider/You.py @@ -19,7 +19,7 @@ class You(BaseProvider): stream: bool, **kwargs: Any, ) -> CreateResult: - url_param = _create_url_param(messages) + url_param = _create_url_param(messages, kwargs.get("history", [])) headers = _create_header() url = f"https://you.com/api/streamingSearch?{url_param}" response = requests.get( @@ -28,12 +28,15 @@ class You(BaseProvider): impersonate="chrome107", ) response.raise_for_status() - yield _parse_output(response.text) + yield _parse_output(response.text).encode().decode("unicode_escape") -def _create_url_param(messages: list[dict[str, str]]): - prompt = messages.pop()["content"] - chat = _convert_chat(messages) +def _create_url_param(messages: list[dict[str, str]], history: list[dict[str, str]]): + prompt = "" + for message in messages: + prompt += "%s: %s\n" % (message["role"], message["content"]) + prompt += "assistant:" + chat = _convert_chat(history) param = {"q": prompt, "domain": "youchat", "chat": chat} return urllib.parse.urlencode(param) -- cgit v1.2.3 From 6a4463889eab6cfa4d73efd2aff7c2904640444c Mon Sep 17 00:00:00 2001 From: zengrr <47846202+zeng-rr@users.noreply.github.com> Date: Wed, 23 Aug 2023 15:34:38 +0800 Subject: fix Yqcloud's prompter using only the last message --- g4f/Provider/Yqcloud.py | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) (limited to 'g4f') diff --git a/g4f/Provider/Yqcloud.py b/g4f/Provider/Yqcloud.py index a3147c2d..82ece2d5 100644 --- a/g4f/Provider/Yqcloud.py +++ b/g4f/Provider/Yqcloud.py @@ -35,7 +35,10 @@ def _create_header(): def _create_payload(messages: list[dict[str, str]]): - prompt = messages[-1]["content"] + prompt = "" + for message in messages: + prompt += "%s: %s\n" % (message["role"], message["content"]) + prompt += "assistant:" return { "prompt": prompt, "network": True, -- cgit v1.2.3 From 1ed448e584092e151212b1d511fd2cbec133064b Mon Sep 17 00:00:00 2001 From: zengrr <47846202+zeng-rr@users.noreply.github.com> Date: Wed, 23 Aug 2023 19:36:30 +0800 Subject: Ails throw the contents of an error response --- g4f/Provider/Ails.py | 2 ++ 1 file changed, 2 insertions(+) (limited to 'g4f') diff --git a/g4f/Provider/Ails.py b/g4f/Provider/Ails.py index 52b3745d..806544b4 100644 --- a/g4f/Provider/Ails.py +++ b/g4f/Provider/Ails.py @@ -72,6 +72,8 @@ class Ails(BaseProvider): if b"content" in token: completion_chunk = json.loads(token.decode().replace("data: ", "")) token = completion_chunk["choices"][0]["delta"].get("content") + if "ai.ls".lower() in token.lower(): + raise Exception("Response Error: " + token) if token != None: yield token -- cgit v1.2.3 From e56f9b7c0dee64eec3843da1fd68f2d0c90b875d Mon Sep 17 00:00:00 2001 From: Bagus Indrayana Date: Thu, 24 Aug 2023 18:13:20 +0800 Subject: remove whitspace from data: and raise exception --- g4f/Provider/EasyChat.py | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-) (limited to 'g4f') diff --git a/g4f/Provider/EasyChat.py b/g4f/Provider/EasyChat.py index 2a61346c..3c6562f0 100644 --- a/g4f/Provider/EasyChat.py +++ b/g4f/Provider/EasyChat.py @@ -75,16 +75,18 @@ class EasyChat(BaseProvider): if "choices" in json_data: yield json_data["choices"][0]["message"]["content"] else: - yield Exception("No response from server") + raise Exception("No response from server") else: for chunk in response.iter_lines(): if b"content" in chunk: - splitData = chunk.decode().split("data: ") + splitData = chunk.decode().split("data:") if len(splitData) > 1: yield json.loads(splitData[1])["choices"][0]["delta"]["content"] + else: + continue else: - yield Exception(f"Error {response.status_code} from server") + raise Exception(f"Error {response.status_code} from server : {response.reason}") @classmethod -- cgit v1.2.3 From 38abb0a0b0a7f2e92dd5d19cee5374cc5a166f20 Mon Sep 17 00:00:00 2001 From: zengrr <47846202+zeng-rr@users.noreply.github.com> Date: Fri, 25 Aug 2023 10:07:48 +0800 Subject: Ails Supplementary Identification of Unusual Response Content --- g4f/Provider/Ails.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'g4f') diff --git a/g4f/Provider/Ails.py b/g4f/Provider/Ails.py index 806544b4..6665b677 100644 --- a/g4f/Provider/Ails.py +++ b/g4f/Provider/Ails.py @@ -72,7 +72,7 @@ class Ails(BaseProvider): if b"content" in token: completion_chunk = json.loads(token.decode().replace("data: ", "")) token = completion_chunk["choices"][0]["delta"].get("content") - if "ai.ls".lower() in token.lower(): + if "ai.ls" in token.lower() or "ai.ci" in token.lower(): raise Exception("Response Error: " + token) if token != None: yield token -- cgit v1.2.3 From 126496d3cacd06a4fa8cbb4e5bde417ce6bb5b4a Mon Sep 17 00:00:00 2001 From: Heiner Lohaus Date: Fri, 25 Aug 2023 06:41:32 +0200 Subject: Add OpenaiChat and Hugchat Provider Add tests for providers with auth Improve async support / 2x faster Shared get_cookies by domain function --- g4f/Provider/Bard.py | 22 ++--------- g4f/Provider/Bing.py | 86 ++++++++++++++----------------------------- g4f/Provider/Hugchat.py | 67 +++++++++++++++++++++++++++++++++ g4f/Provider/OpenaiChat.py | 74 +++++++++++++++++++++++++++++++++++++ g4f/Provider/__init__.py | 4 ++ g4f/Provider/base_provider.py | 85 +++++++++++++++++++++++++++++++++++++++++- 6 files changed, 259 insertions(+), 79 deletions(-) create mode 100644 g4f/Provider/Hugchat.py create mode 100644 g4f/Provider/OpenaiChat.py (limited to 'g4f') diff --git a/g4f/Provider/Bard.py b/g4f/Provider/Bard.py index cbe728cd..a8c7d13f 100644 --- a/g4f/Provider/Bard.py +++ b/g4f/Provider/Bard.py @@ -2,42 +2,26 @@ import json import random import re -import browser_cookie3 from aiohttp import ClientSession import asyncio from ..typing import Any, CreateResult -from .base_provider import BaseProvider +from .base_provider import AsyncProvider, get_cookies -class Bard(BaseProvider): +class Bard(AsyncProvider): url = "https://bard.google.com" needs_auth = True working = True - @classmethod - def create_completion( - cls, - model: str, - messages: list[dict[str, str]], - stream: bool, - proxy: str = None, - cookies: dict = {}, - **kwargs: Any, - ) -> CreateResult: - yield asyncio.run(cls.create_async(str, messages, proxy, cookies)) - @classmethod async def create_async( cls, model: str, messages: list[dict[str, str]], proxy: str = None, - cookies: dict = {}, + cookies: dict = get_cookies(".google.com"), **kwargs: Any, ) -> str: - if not cookies: - for cookie in browser_cookie3.load(domain_name='.google.com'): - cookies[cookie.name] = cookie.value formatted = "\n".join( ["%s: %s" % (message["role"], message["content"]) for message in messages] diff --git a/g4f/Provider/Bing.py b/g4f/Provider/Bing.py index b9b9e9a4..2c2e60ad 100644 --- a/g4f/Provider/Bing.py +++ b/g4f/Provider/Bing.py @@ -5,48 +5,24 @@ import random import aiohttp import asyncio -import browser_cookie3 from aiohttp import ClientSession from ..typing import Any, AsyncGenerator, CreateResult, Union -from .base_provider import BaseProvider +from .base_provider import AsyncGeneratorProvider, get_cookies -class Bing(BaseProvider): +class Bing(AsyncGeneratorProvider): url = "https://bing.com/chat" + needs_auth = True working = True supports_gpt_4 = True - - @classmethod - def create_completion( - cls, - model: str, - messages: list[dict[str, str]], - stream: bool, - **kwargs: Any - ) -> CreateResult: - if stream: - yield from run(cls.create_async_generator(model, messages, **kwargs)) - else: - yield asyncio.run(cls.create_async(model, messages, **kwargs)) - - @classmethod - async def create_async( - cls, - model: str, - messages: list[dict[str, str]], - **kwargs: Any, - ) -> str: - result = [] - async for chunk in cls.create_async_generator(model, messages, **kwargs): - result.append(chunk) - if result: - return "".join(result) + supports_stream=True @staticmethod def create_async_generator( model: str, messages: list[dict[str, str]], - cookies: dict = {} + cookies: dict = get_cookies(".bing.com"), + **kwargs ) -> AsyncGenerator: if len(messages) < 2: prompt = messages[0]["content"] @@ -54,15 +30,11 @@ class Bing(BaseProvider): else: prompt = messages[-1]["content"] - context = convert(messages[:-1]) - - if not cookies: - for cookie in browser_cookie3.load(domain_name='.bing.com'): - cookies[cookie.name] = cookie.value + context = create_context(messages[:-1]) return stream_generate(prompt, context, cookies) -def convert(messages: list[dict[str, str]]): +def create_context(messages: list[dict[str, str]]): context = "" for message in messages: @@ -187,34 +159,32 @@ class Defaults: 'x-forwarded-for': ip_address, } - optionsSets = { - "optionsSets": [ - 'saharasugg', - 'enablenewsfc', - 'clgalileo', - 'gencontentv3', - "nlu_direct_response_filter", - "deepleo", - "disable_emoji_spoken_text", - "responsible_ai_policy_235", - "enablemm", - "h3precise" - "dtappid", - "cricinfo", - "cricinfov2", - "dv3sugg", - "nojbfedge" - ] - } + optionsSets = [ + 'saharasugg', + 'enablenewsfc', + 'clgalileo', + 'gencontentv3', + "nlu_direct_response_filter", + "deepleo", + "disable_emoji_spoken_text", + "responsible_ai_policy_235", + "enablemm", + "h3precise" + "dtappid", + "cricinfo", + "cricinfov2", + "dv3sugg", + "nojbfedge" + ] -def format_message(msg: dict) -> str: - return json.dumps(msg, ensure_ascii=False) + Defaults.delimiter +def format_message(message: dict) -> str: + return json.dumps(message, ensure_ascii=False) + Defaults.delimiter def create_message(conversation: Conversation, prompt: str, context: str=None) -> str: struct = { 'arguments': [ { - **Defaults.optionsSets, + 'optionsSets': Defaults.optionsSets, 'source': 'cib', 'allowedMessageTypes': Defaults.allowedMessageTypes, 'sliceIds': Defaults.sliceIds, diff --git a/g4f/Provider/Hugchat.py b/g4f/Provider/Hugchat.py new file mode 100644 index 00000000..cedf8402 --- /dev/null +++ b/g4f/Provider/Hugchat.py @@ -0,0 +1,67 @@ +has_module = False +try: + from hugchat.hugchat import ChatBot +except ImportError: + has_module = False + +from .base_provider import BaseProvider, get_cookies +from g4f.typing import CreateResult + +class Hugchat(BaseProvider): + url = "https://huggingface.co/chat/" + needs_auth = True + working = has_module + llms = ['OpenAssistant/oasst-sft-6-llama-30b-xor', 'meta-llama/Llama-2-70b-chat-hf'] + + @classmethod + def create_completion( + cls, + model: str, + messages: list[dict[str, str]], + stream: bool = False, + proxy: str = None, + cookies: str = get_cookies(".huggingface.co"), + **kwargs + ) -> CreateResult: + bot = ChatBot( + cookies=cookies + ) + + if proxy and "://" not in proxy: + proxy = f"http://{proxy}" + bot.session.proxies = {"http": proxy, "https": proxy} + + if model: + try: + if not isinstance(model, int): + model = cls.llms.index(model) + bot.switch_llm(model) + except: + raise RuntimeError(f"Model are not supported: {model}") + + if len(messages) > 1: + formatted = "\n".join( + ["%s: %s" % (message["role"], message["content"]) for message in messages] + ) + prompt = f"{formatted}\nAssistant:" + else: + prompt = messages.pop()["content"] + + try: + yield bot.chat(prompt, **kwargs) + finally: + bot.delete_conversation(bot.current_conversation) + bot.current_conversation = "" + pass + + @classmethod + @property + def params(cls): + params = [ + ("model", "str"), + ("messages", "list[dict[str, str]]"), + ("stream", "bool"), + ("proxy", "str"), + ] + param = ", ".join([": ".join(p) for p in params]) + return f"g4f.provider.{cls.__name__} supports: ({param})" diff --git a/g4f/Provider/OpenaiChat.py b/g4f/Provider/OpenaiChat.py new file mode 100644 index 00000000..cca258b3 --- /dev/null +++ b/g4f/Provider/OpenaiChat.py @@ -0,0 +1,74 @@ +has_module = True +try: + from revChatGPT.V1 import AsyncChatbot +except ImportError: + has_module = False +from .base_provider import AsyncGeneratorProvider, get_cookies +from ..typing import AsyncGenerator + +class OpenaiChat(AsyncGeneratorProvider): + url = "https://chat.openai.com" + needs_auth = True + working = has_module + supports_gpt_35_turbo = True + supports_gpt_4 = True + supports_stream = True + + @classmethod + async def create_async_generator( + cls, + model: str, + messages: list[dict[str, str]], + proxy: str = None, + access_token: str = None, + cookies: dict = None, + **kwargs + ) -> AsyncGenerator: + + config = {"access_token": access_token, "model": model} + if proxy: + if "://" not in proxy: + proxy = f"http://{proxy}" + config["proxy"] = proxy + + bot = AsyncChatbot( + config=config + ) + + if not access_token: + cookies = cookies if cookies else get_cookies("chat.openai.com") + response = await bot.session.get("https://chat.openai.com/api/auth/session", cookies=cookies) + access_token = response.json()["accessToken"] + bot.set_access_token(access_token) + + if len(messages) > 1: + formatted = "\n".join( + ["%s: %s" % ((message["role"]).capitalize(), message["content"]) for message in messages] + ) + prompt = f"{formatted}\nAssistant:" + else: + prompt = messages.pop()["content"] + + returned = None + async for message in bot.ask(prompt): + message = message["message"] + if returned: + if message.startswith(returned): + new = message[len(returned):] + if new: + yield new + else: + yield message + returned = message + + @classmethod + @property + def params(cls): + params = [ + ("model", "str"), + ("messages", "list[dict[str, str]]"), + ("stream", "bool"), + ("proxy", "str"), + ] + param = ", ".join([": ".join(p) for p in params]) + return f"g4f.provider.{cls.__name__} supports: ({param})" diff --git a/g4f/Provider/__init__.py b/g4f/Provider/__init__.py index e27dee5d..5ad9f156 100644 --- a/g4f/Provider/__init__.py +++ b/g4f/Provider/__init__.py @@ -14,9 +14,11 @@ from .EasyChat import EasyChat from .Forefront import Forefront from .GetGpt import GetGpt from .H2o import H2o +from .Hugchat import Hugchat from .Liaobots import Liaobots from .Lockchat import Lockchat from .Opchatgpts import Opchatgpts +from .OpenaiChat import OpenaiChat from .Raycast import Raycast from .Theb import Theb from .Vercel import Vercel @@ -44,10 +46,12 @@ __all__ = [ "Forefront", "GetGpt", "H2o", + "Hugchat", "Liaobots", "Lockchat", "Opchatgpts", "Raycast", + "OpenaiChat", "Theb", "Vercel", "Wewordle", diff --git a/g4f/Provider/base_provider.py b/g4f/Provider/base_provider.py index 98ad3514..56d79ee6 100644 --- a/g4f/Provider/base_provider.py +++ b/g4f/Provider/base_provider.py @@ -1,7 +1,11 @@ from abc import ABC, abstractmethod -from ..typing import Any, CreateResult +from ..typing import Any, CreateResult, AsyncGenerator, Union +import browser_cookie3 +import asyncio +from time import time +import math class BaseProvider(ABC): url: str @@ -30,4 +34,81 @@ class BaseProvider(ABC): ("stream", "bool"), ] param = ", ".join([": ".join(p) for p in params]) - return f"g4f.provider.{cls.__name__} supports: ({param})" \ No newline at end of file + return f"g4f.provider.{cls.__name__} supports: ({param})" + + +_cookies = {} + +def get_cookies(cookie_domain: str) -> dict: + if cookie_domain not in _cookies: + _cookies[cookie_domain] = {} + for cookie in browser_cookie3.load(cookie_domain): + _cookies[cookie_domain][cookie.name] = cookie.value + return _cookies[cookie_domain] + + +class AsyncProvider(BaseProvider): + @classmethod + def create_completion( + cls, + model: str, + messages: list[dict[str, str]], + stream: bool = False, + **kwargs: Any + ) -> CreateResult: + yield asyncio.run(cls.create_async(model, messages, **kwargs)) + + @staticmethod + @abstractmethod + async def create_async( + model: str, + messages: list[dict[str, str]], + **kwargs: Any, + ) -> str: + raise NotImplementedError() + + +class AsyncGeneratorProvider(AsyncProvider): + @classmethod + def create_completion( + cls, + model: str, + messages: list[dict[str, str]], + stream: bool = True, + **kwargs: Any + ) -> CreateResult: + if stream: + yield from run_generator(cls.create_async_generator(model, messages, **kwargs)) + else: + yield from AsyncProvider.create_completion(cls=cls, model=model, messages=messages, **kwargs) + + @classmethod + async def create_async( + cls, + model: str, + messages: list[dict[str, str]], + **kwargs: Any, + ) -> str: + chunks = [chunk async for chunk in cls.create_async_generator(model, messages, **kwargs)] + if chunks: + return "".join(chunks) + + @staticmethod + @abstractmethod + def create_async_generator( + model: str, + messages: list[dict[str, str]], + ) -> AsyncGenerator: + raise NotImplementedError() + + +def run_generator(generator: AsyncGenerator[Union[Any, str], Any]): + loop = asyncio.new_event_loop() + gen = generator.__aiter__() + + while True: + try: + yield loop.run_until_complete(gen.__anext__()) + + except StopAsyncIteration: + break -- cgit v1.2.3