From ec51e9c76433b9e9498ade1dbe5de2268ad84218 Mon Sep 17 00:00:00 2001 From: Heiner Lohaus Date: Mon, 11 Mar 2024 02:41:59 +0100 Subject: Fix HuggingChat and PerplexityLabs and add HuggingFace provider Add more models and image generation in You provider Use You as second default image provider --- g4f/Provider/HuggingChat.py | 35 +++++++++++++++++++++-------------- 1 file changed, 21 insertions(+), 14 deletions(-) (limited to 'g4f/Provider/HuggingChat.py') diff --git a/g4f/Provider/HuggingChat.py b/g4f/Provider/HuggingChat.py index 3ea9f306..9644880c 100644 --- a/g4f/Provider/HuggingChat.py +++ b/g4f/Provider/HuggingChat.py @@ -1,12 +1,12 @@ from __future__ import annotations -import json, uuid +import json from aiohttp import ClientSession, BaseConnector from ..typing import AsyncResult, Messages from .base_provider import AsyncGeneratorProvider, ProviderModelMixin -from .helper import format_prompt, get_cookies, get_connector +from .helper import format_prompt, get_connector class HuggingChat(AsyncGeneratorProvider, ProviderModelMixin): @@ -24,7 +24,6 @@ class HuggingChat(AsyncGeneratorProvider, ProviderModelMixin): ] model_aliases = { "openchat/openchat_3.5": "openchat/openchat-3.5-1210", - "mistralai/Mixtral-8x7B-Instruct-v0.1": "mistralai/Mistral-7B-Instruct-v0.2" } @classmethod @@ -39,9 +38,11 @@ class HuggingChat(AsyncGeneratorProvider, ProviderModelMixin): cookies: dict = None, **kwargs ) -> AsyncResult: - if not cookies: - cookies = get_cookies(".huggingface.co", False) - + options = {"model": cls.get_model(model)} + system_prompt = "\n".join([message["content"] for message in messages if message["role"] == "system"]) + if system_prompt: + options["preprompt"] = system_prompt + messages = [message for message in messages if message["role"] != "system"] headers = { 'User-Agent': 'Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/111.0.0.0 Safari/537.36', } @@ -50,20 +51,27 @@ class HuggingChat(AsyncGeneratorProvider, ProviderModelMixin): headers=headers, connector=get_connector(connector, proxy) ) as session: - async with session.post(f"{cls.url}/conversation", json={"model": cls.get_model(model)}, proxy=proxy) as response: + async with session.post(f"{cls.url}/conversation", json=options, proxy=proxy) as response: + response.raise_for_status() conversation_id = (await response.json())["conversationId"] - - send = { - "id": str(uuid.uuid4()), + async with session.get(f"{cls.url}/conversation/{conversation_id}/__data.json") as response: + response.raise_for_status() + data: list = (await response.json())["nodes"][1]["data"] + keys: list[int] = data[data[0]["messages"]] + message_keys: dict = data[keys[0]] + message_id: str = data[message_keys["id"]] + options = { + "id": message_id, "inputs": format_prompt(messages), + "is_continue": False, "is_retry": False, - "response_id": str(uuid.uuid4()), "web_search": web_search } - async with session.post(f"{cls.url}/conversation/{conversation_id}", json=send, proxy=proxy) as response: + async with session.post(f"{cls.url}/conversation/{conversation_id}", json=options) as response: first_token = True async for line in response.content: - line = json.loads(line[:-1]) + response.raise_for_status() + line = json.loads(line) if "type" not in line: raise RuntimeError(f"Response: {line}") elif line["type"] == "stream": @@ -74,6 +82,5 @@ class HuggingChat(AsyncGeneratorProvider, ProviderModelMixin): yield token elif line["type"] == "finalAnswer": break - async with session.delete(f"{cls.url}/conversation/{conversation_id}", proxy=proxy) as response: response.raise_for_status() -- cgit v1.2.3