diff options
author | H Lohaus <hlohaus@users.noreply.github.com> | 2024-03-26 21:50:35 +0100 |
---|---|---|
committer | GitHub <noreply@github.com> | 2024-03-26 21:50:35 +0100 |
commit | dd08125bb4740ab5dc1116868ff237ce51ae3d5e (patch) | |
tree | ed6339c66505409e7dced6d3499814b00446ff12 /g4f/Provider/needs_auth | |
parent | Merge pull request #1754 from hlohaus/goo (diff) | |
parent | Fix load .har files, add hardir to docker, add docs (diff) | |
download | gpt4free-dd08125bb4740ab5dc1116868ff237ce51ae3d5e.tar gpt4free-dd08125bb4740ab5dc1116868ff237ce51ae3d5e.tar.gz gpt4free-dd08125bb4740ab5dc1116868ff237ce51ae3d5e.tar.bz2 gpt4free-dd08125bb4740ab5dc1116868ff237ce51ae3d5e.tar.lz gpt4free-dd08125bb4740ab5dc1116868ff237ce51ae3d5e.tar.xz gpt4free-dd08125bb4740ab5dc1116868ff237ce51ae3d5e.tar.zst gpt4free-dd08125bb4740ab5dc1116868ff237ce51ae3d5e.zip |
Diffstat (limited to 'g4f/Provider/needs_auth')
-rw-r--r-- | g4f/Provider/needs_auth/OpenaiChat.py | 50 |
1 files changed, 21 insertions, 29 deletions
diff --git a/g4f/Provider/needs_auth/OpenaiChat.py b/g4f/Provider/needs_auth/OpenaiChat.py index 0aff99a7..72f9f224 100644 --- a/g4f/Provider/needs_auth/OpenaiChat.py +++ b/g4f/Provider/needs_auth/OpenaiChat.py @@ -3,10 +3,10 @@ from __future__ import annotations import asyncio import uuid import json -import os import base64 import time from aiohttp import ClientWebSocketResponse +from copy import copy try: import webview @@ -22,13 +22,13 @@ except ImportError: pass from ..base_provider import AsyncGeneratorProvider, ProviderModelMixin -from ..helper import get_cookies from ...webdriver import get_browser from ...typing import AsyncResult, Messages, Cookies, ImageType, Union, AsyncIterator from ...requests import get_args_from_browser, raise_for_status from ...requests.aiohttp import StreamSession from ...image import to_image, to_bytes, ImageResponse, ImageRequest -from ...errors import MissingRequirementsError, MissingAuthError, ProviderNotWorkingError +from ...errors import MissingAuthError +from ...providers.conversation import BaseConversation from ..openai.har_file import getArkoseAndAccessToken from ... import debug @@ -56,11 +56,6 @@ class OpenaiChat(AsyncGeneratorProvider, ProviderModelMixin): prompt: str = None, model: str = "", messages: Messages = [], - history_disabled: bool = False, - action: str = "next", - conversation_id: str = None, - parent_id: str = None, - image: ImageType = None, **kwargs ) -> Response: """ @@ -89,12 +84,7 @@ class OpenaiChat(AsyncGeneratorProvider, ProviderModelMixin): generator = cls.create_async_generator( model, messages, - history_disabled=history_disabled, - action=action, - conversation_id=conversation_id, - parent_id=parent_id, - image=image, - response_fields=True, + return_conversation=True, **kwargs ) return Response( @@ -209,7 +199,7 @@ class OpenaiChat(AsyncGeneratorProvider, ProviderModelMixin): } for message in messages] # Check if there is an image response - if image_request: + if image_request is not None: # Change content in last user message messages[-1]["content"] = { "content_type": "multimodal_text", @@ -308,10 +298,11 @@ class OpenaiChat(AsyncGeneratorProvider, ProviderModelMixin): history_disabled: bool = True, action: str = "next", conversation_id: str = None, + conversation: Conversation = None, parent_id: str = None, image: ImageType = None, image_name: str = None, - response_fields: bool = False, + return_conversation: bool = False, **kwargs ) -> AsyncResult: """ @@ -330,7 +321,7 @@ class OpenaiChat(AsyncGeneratorProvider, ProviderModelMixin): conversation_id (str): ID of the conversation. parent_id (str): ID of the parent message. image (ImageType): Image to include in the conversation. - response_fields (bool): Flag to include response fields in the output. + return_conversation (bool): Flag to include response fields in the output. **kwargs: Additional keyword arguments. Yields: @@ -387,6 +378,8 @@ class OpenaiChat(AsyncGeneratorProvider, ProviderModelMixin): arkose_token, api_key, cookies = await getArkoseAndAccessToken(proxy) cls._create_request_args(cookies) cls._set_api_key(api_key) + if arkose_token is None: + raise MissingAuthError("No arkose token found in .har file") try: image_request = await cls.upload_image(session, cls._headers, image, image_name) if image else None @@ -396,7 +389,8 @@ class OpenaiChat(AsyncGeneratorProvider, ProviderModelMixin): print(f"{e.__class__.__name__}: {e}") model = cls.get_model(model).replace("gpt-3.5-turbo", "text-davinci-002-render-sha") - fields = ResponseFields() + fields = Conversation() if conversation is None else copy(conversation) + fields.finish_reason = None while fields.finish_reason is None: conversation_id = conversation_id if fields.conversation_id is None else fields.conversation_id parent_id = parent_id if fields.message_id is None else fields.message_id @@ -409,7 +403,7 @@ class OpenaiChat(AsyncGeneratorProvider, ProviderModelMixin): "conversation_id": conversation_id, "parent_message_id": parent_id, "model": model, - "history_and_training_disabled": history_disabled and not auto_continue, + "history_and_training_disabled": history_disabled and not auto_continue and not return_conversation, "websocket_request_id": websocket_request_id } if action != "continue": @@ -422,8 +416,6 @@ class OpenaiChat(AsyncGeneratorProvider, ProviderModelMixin): } if need_arkose: headers["OpenAI-Sentinel-Arkose-Token"] = arkose_token - headers["OpenAI-Sentinel-Chat-Requirements-Token"] = chat_token - async with session.post( f"{cls.url}/backend-api/conversation", json=data, @@ -432,15 +424,15 @@ class OpenaiChat(AsyncGeneratorProvider, ProviderModelMixin): cls._update_request_args(session) await raise_for_status(response) async for chunk in cls.iter_messages_chunk(response.iter_lines(), session, fields): - if response_fields: - response_fields = False + if return_conversation: + return_conversation = False yield fields yield chunk if not auto_continue: break action = "continue" await asyncio.sleep(5) - if history_disabled and auto_continue: + if history_disabled and auto_continue and not return_conversation: await cls.delete_conversation(session, cls._headers, fields.conversation_id) @staticmethod @@ -458,7 +450,7 @@ class OpenaiChat(AsyncGeneratorProvider, ProviderModelMixin): cls, messages: AsyncIterator, session: StreamSession, - fields: ResponseFields + fields: Conversation ) -> AsyncIterator: last_message: int = 0 async for message in messages: @@ -487,7 +479,7 @@ class OpenaiChat(AsyncGeneratorProvider, ProviderModelMixin): break @classmethod - async def iter_messages_line(cls, session: StreamSession, line: bytes, fields: ResponseFields) -> AsyncIterator: + async def iter_messages_line(cls, session: StreamSession, line: bytes, fields: Conversation) -> AsyncIterator: if not line.startswith(b"data: "): return elif line.startswith(b"data: [DONE]"): @@ -618,7 +610,7 @@ this.fetch = async (url, options) => { @classmethod def _update_request_args(cls, session: StreamSession): for c in session.cookie_jar if hasattr(session, "cookie_jar") else session.cookies.jar: - cls._cookies[c.name if hasattr(c, "name") else c.key] = c.value + cls._cookies[c.key if hasattr(c, "key") else c.name] = c.value cls._update_cookie_header() @classmethod @@ -631,7 +623,7 @@ this.fetch = async (url, options) => { def _update_cookie_header(cls): cls._headers["Cookie"] = cls._format_cookies(cls._cookies) -class ResponseFields: +class Conversation(BaseConversation): """ Class to encapsulate response fields. """ @@ -664,7 +656,7 @@ class Response(): self._generator = None chunks = [] async for chunk in self._generator: - if isinstance(chunk, ResponseFields): + if isinstance(chunk, Conversation): self._fields = chunk else: yield chunk |