mirror of
https://github.com/xtekky/gpt4free.git
synced 2024-11-24 01:36:37 +03:00
42 lines
1.2 KiB
Python
42 lines
1.2 KiB
Python
import json
|
|
|
|
import requests
|
|
|
|
from ..typing import Any, CreateResult
|
|
from .base_provider import BaseProvider
|
|
|
|
|
|
class Forefront(BaseProvider):
|
|
url = "https://forefront.com"
|
|
supports_stream = True
|
|
supports_gpt_35_turbo = True
|
|
|
|
@staticmethod
|
|
def create_completion(
|
|
model: str,
|
|
messages: list[dict[str, str]],
|
|
stream: bool,
|
|
**kwargs: Any,
|
|
) -> CreateResult:
|
|
json_data = {
|
|
"text": messages[-1]["content"],
|
|
"action": "noauth",
|
|
"id": "",
|
|
"parentId": "",
|
|
"workspaceId": "",
|
|
"messagePersona": "607e41fe-95be-497e-8e97-010a59b2e2c0",
|
|
"model": "gpt-4",
|
|
"messages": messages[:-1] if len(messages) > 1 else [],
|
|
"internetMode": "auto",
|
|
}
|
|
|
|
response = requests.post(
|
|
"https://streaming.tenant-forefront-default.knative.chi.coreweave.com/free-chat",
|
|
json=json_data,
|
|
stream=True,
|
|
)
|
|
response.raise_for_status()
|
|
for token in response.iter_lines():
|
|
if b"delta" in token:
|
|
yield json.loads(token.decode().split("data: ")[1])["delta"]
|