|
from __future__ import annotations |
|
|
|
import json |
|
|
|
import requests |
|
|
|
from ..typing import Any, CreateResult |
|
from .base_provider import BaseProvider |
|
|
|
|
|
class Raycast(BaseProvider): |
|
url = "https://raycast.com" |
|
supports_gpt_35_turbo = True |
|
supports_gpt_4 = True |
|
supports_stream = True |
|
needs_auth = True |
|
working = True |
|
|
|
@staticmethod |
|
def create_completion( |
|
model: str, |
|
messages: list[dict[str, str]], |
|
stream: bool, |
|
**kwargs: Any, |
|
) -> CreateResult: |
|
auth = kwargs.get('auth') |
|
headers = { |
|
'Accept': 'application/json', |
|
'Accept-Language': 'en-US,en;q=0.9', |
|
'Authorization': f'Bearer {auth}', |
|
'Content-Type': 'application/json', |
|
'User-Agent': 'Raycast/0 CFNetwork/1410.0.3 Darwin/22.6.0', |
|
} |
|
parsed_messages = [] |
|
for message in messages: |
|
parsed_messages.append({ |
|
'author': message['role'], |
|
'content': {'text': message['content']} |
|
}) |
|
data = { |
|
"debug": False, |
|
"locale": "en-CN", |
|
"messages": parsed_messages, |
|
"model": model, |
|
"provider": "openai", |
|
"source": "ai_chat", |
|
"system_instruction": "markdown", |
|
"temperature": 0.5 |
|
} |
|
response = requests.post("https://backend.raycast.com/api/v1/ai/chat_completions", headers=headers, json=data, stream=True) |
|
for token in response.iter_lines(): |
|
if b'data: ' not in token: |
|
continue |
|
completion_chunk = json.loads(token.decode().replace('data: ', '')) |
|
token = completion_chunk['text'] |
|
if token != None: |
|
yield token |
|
|
|
@classmethod |
|
@property |
|
def params(cls): |
|
params = [ |
|
("model", "str"), |
|
("messages", "list[dict[str, str]]"), |
|
("stream", "bool"), |
|
("temperature", "float"), |
|
("top_p", "int"), |
|
("model", "str"), |
|
("auth", "str"), |
|
] |
|
param = ", ".join([": ".join(p) for p in params]) |
|
return f"g4f.provider.{cls.__name__} supports: ({param})" |
|
|