mirror of
https://github.com/NovaOSS/nova-api.git
synced 2024-11-25 22:43:57 +01:00
Compare commits
1 commit
1efb527961
...
ec31a268ee
Author | SHA1 | Date | |
---|---|---|---|
ec31a268ee |
|
@ -1,17 +1,21 @@
|
||||||
import os
|
import os
|
||||||
import time
|
import time
|
||||||
import asyncio
|
|
||||||
|
|
||||||
from dotenv import load_dotenv
|
from dotenv import load_dotenv
|
||||||
from motor.motor_asyncio import AsyncIOMotorClient
|
from motor.motor_asyncio import AsyncIOMotorClient
|
||||||
|
|
||||||
try:
|
from helpers import network
|
||||||
from helpers import network
|
|
||||||
except ImportError:
|
|
||||||
pass
|
|
||||||
|
|
||||||
load_dotenv()
|
load_dotenv()
|
||||||
|
|
||||||
|
UA_SIMPLIFY = {
|
||||||
|
'Windows NT': 'W',
|
||||||
|
'Mozilla/5.0': 'M',
|
||||||
|
'Win64; x64': '64',
|
||||||
|
'Safari/537.36': 'S',
|
||||||
|
'AppleWebKit/537.36 (KHTML, like Gecko)': 'K',
|
||||||
|
}
|
||||||
|
|
||||||
## MONGODB Setup
|
## MONGODB Setup
|
||||||
|
|
||||||
conn = AsyncIOMotorClient(os.environ['MONGO_URI'])
|
conn = AsyncIOMotorClient(os.environ['MONGO_URI'])
|
||||||
|
@ -26,7 +30,18 @@ async def replacer(text: str, dict_: dict) -> str:
|
||||||
return text
|
return text
|
||||||
|
|
||||||
async def log_api_request(user: dict, incoming_request, target_url: str):
|
async def log_api_request(user: dict, incoming_request, target_url: str):
|
||||||
"""Logs the API Request into the database."""
|
"""Logs the API Request into the database.
|
||||||
|
No input prompt is logged, however data such as IP & useragent is noted.
|
||||||
|
This would be useful for security reasons. Other minor data is also collected.
|
||||||
|
|
||||||
|
Args:
|
||||||
|
user (dict): User dict object
|
||||||
|
incoming_request (_type_): Request
|
||||||
|
target_url (str): The URL the api request was targetted to.
|
||||||
|
|
||||||
|
Returns:
|
||||||
|
_type_: _description_
|
||||||
|
"""
|
||||||
db = await _get_collection('logs')
|
db = await _get_collection('logs')
|
||||||
payload = {}
|
payload = {}
|
||||||
|
|
||||||
|
@ -38,6 +53,7 @@ async def log_api_request(user: dict, incoming_request, target_url: str):
|
||||||
|
|
||||||
model = payload.get('model')
|
model = payload.get('model')
|
||||||
ip_address = await network.get_ip(incoming_request)
|
ip_address = await network.get_ip(incoming_request)
|
||||||
|
useragent = await replacer(incoming_request.headers.get('User-Agent', ''), UA_SIMPLIFY)
|
||||||
|
|
||||||
new_log_item = {
|
new_log_item = {
|
||||||
'timestamp': time.time(),
|
'timestamp': time.time(),
|
||||||
|
@ -46,6 +62,7 @@ async def log_api_request(user: dict, incoming_request, target_url: str):
|
||||||
'user_id': str(user['_id']),
|
'user_id': str(user['_id']),
|
||||||
'security': {
|
'security': {
|
||||||
'ip': ip_address,
|
'ip': ip_address,
|
||||||
|
'useragent': useragent,
|
||||||
},
|
},
|
||||||
'details': {
|
'details': {
|
||||||
'model': model,
|
'model': model,
|
||||||
|
@ -73,21 +90,5 @@ async def delete_by_user_id(user_id: str):
|
||||||
db = await _get_collection('logs')
|
db = await _get_collection('logs')
|
||||||
return await db.delete_many({'user_id': user_id})
|
return await db.delete_many({'user_id': user_id})
|
||||||
|
|
||||||
async def get_logs_time_range(start: int, end: int):
|
|
||||||
db = await _get_collection('logs')
|
|
||||||
|
|
||||||
entries = []
|
|
||||||
async for entry in db.find({'timestamp': {'$gte': start, '$lte': end}}):
|
|
||||||
entries.append(entry)
|
|
||||||
|
|
||||||
return entries
|
|
||||||
|
|
||||||
async def main():
|
|
||||||
# how many requests in last 24 hours?
|
|
||||||
last_24_hours = time.time() - 86400
|
|
||||||
logs = await get_logs_time_range(last_24_hours, time.time())
|
|
||||||
|
|
||||||
print(f'Number of logs in last 24 hours: {len(logs)}')
|
|
||||||
|
|
||||||
if __name__ == '__main__':
|
if __name__ == '__main__':
|
||||||
asyncio.run(main())
|
pass
|
||||||
|
|
|
@ -100,7 +100,7 @@ manager = KeyManager()
|
||||||
|
|
||||||
async def main():
|
async def main():
|
||||||
keys = await manager.get_possible_keys('closed')
|
keys = await manager.get_possible_keys('closed')
|
||||||
print(keys)
|
print(len(keys))
|
||||||
|
|
||||||
if __name__ == '__main__':
|
if __name__ == '__main__':
|
||||||
asyncio.run(main())
|
asyncio.run(main())
|
||||||
|
|
|
@ -2,6 +2,8 @@ import os
|
||||||
import pytz
|
import pytz
|
||||||
import asyncio
|
import asyncio
|
||||||
import datetime
|
import datetime
|
||||||
|
import json
|
||||||
|
import time
|
||||||
|
|
||||||
from dotenv import load_dotenv
|
from dotenv import load_dotenv
|
||||||
from motor.motor_asyncio import AsyncIOMotorClient
|
from motor.motor_asyncio import AsyncIOMotorClient
|
||||||
|
@ -13,6 +15,13 @@ load_dotenv()
|
||||||
class StatsManager:
|
class StatsManager:
|
||||||
"""
|
"""
|
||||||
### The manager for all statistics tracking
|
### The manager for all statistics tracking
|
||||||
|
Stats tracked:
|
||||||
|
- Dates
|
||||||
|
- IPs
|
||||||
|
- Target URLs
|
||||||
|
- Tokens
|
||||||
|
- Models
|
||||||
|
- URL Paths
|
||||||
"""
|
"""
|
||||||
|
|
||||||
def __init__(self):
|
def __init__(self):
|
||||||
|
|
6
api/db/tester.py
Normal file
6
api/db/tester.py
Normal file
|
@ -0,0 +1,6 @@
|
||||||
|
from stats import *
|
||||||
|
import asyncio
|
||||||
|
|
||||||
|
manager = StatsManager()
|
||||||
|
|
||||||
|
asyncio.run(manager.get_model_usage())
|
|
@ -1,12 +1,12 @@
|
||||||
from . import \
|
from . import \
|
||||||
azure \
|
azure, \
|
||||||
# closed, \
|
closed, \
|
||||||
# closed4
|
closed4
|
||||||
# closed432
|
# closed432
|
||||||
|
|
||||||
MODULES = [
|
MODULES = [
|
||||||
azure,
|
azure,
|
||||||
# closed,
|
closed,
|
||||||
# closed4,
|
closed4,
|
||||||
# closed432,
|
# closed432,
|
||||||
]
|
]
|
||||||
|
|
|
@ -12,7 +12,7 @@ MODELS = [
|
||||||
'gpt-4',
|
'gpt-4',
|
||||||
'gpt-4-32k'
|
'gpt-4-32k'
|
||||||
]
|
]
|
||||||
# MODELS = [f'{model}-azure' for model in MODELS]
|
MODELS = [f'{model}-azure' for model in MODELS]
|
||||||
|
|
||||||
AZURE_API = '2023-07-01-preview'
|
AZURE_API = '2023-07-01-preview'
|
||||||
|
|
||||||
|
|
|
@ -34,4 +34,7 @@ async def conversation_to_prompt(conversation: list) -> str:
|
||||||
return text
|
return text
|
||||||
|
|
||||||
async def random_secret_for(name: str) -> str:
|
async def random_secret_for(name: str) -> str:
|
||||||
return await providerkeys.manager.get_key(name)
|
try:
|
||||||
|
return await providerkeys.manager.get_key(name)
|
||||||
|
except ValueError:
|
||||||
|
raise ValueError(f'Keys missing for "{name}" <no_keys>')
|
||||||
|
|
|
@ -2,7 +2,7 @@
|
||||||
|
|
||||||
import os
|
import os
|
||||||
import json
|
import json
|
||||||
import ujson
|
import logging
|
||||||
import aiohttp
|
import aiohttp
|
||||||
import asyncio
|
import asyncio
|
||||||
import starlette
|
import starlette
|
||||||
|
@ -49,7 +49,9 @@ async def respond(
|
||||||
'Content-Type': 'application/json'
|
'Content-Type': 'application/json'
|
||||||
}
|
}
|
||||||
|
|
||||||
for i in range(1):
|
for i in range(20):
|
||||||
|
print(i)
|
||||||
|
# Load balancing: randomly selecting a suitable provider
|
||||||
try:
|
try:
|
||||||
if is_chat:
|
if is_chat:
|
||||||
target_request = await load_balancing.balance_chat_request(payload)
|
target_request = await load_balancing.balance_chat_request(payload)
|
||||||
|
@ -149,21 +151,13 @@ async def respond(
|
||||||
|
|
||||||
async for chunk in response.content.iter_any():
|
async for chunk in response.content.iter_any():
|
||||||
chunk = chunk.decode('utf8').strip()
|
chunk = chunk.decode('utf8').strip()
|
||||||
|
|
||||||
if 'azure' in provider_name:
|
|
||||||
chunk = chunk.strip().replace('data: ', '')
|
|
||||||
|
|
||||||
if not chunk or 'prompt_filter_results' in chunk:
|
|
||||||
continue
|
|
||||||
|
|
||||||
yield chunk + '\n\n'
|
yield chunk + '\n\n'
|
||||||
|
|
||||||
break
|
break
|
||||||
|
|
||||||
except Exception as exc:
|
except Exception as exc:
|
||||||
print('[!] exception', exc)
|
print('[!] exception', exc)
|
||||||
# continue
|
continue
|
||||||
raise exc
|
|
||||||
|
|
||||||
else:
|
else:
|
||||||
yield await errors.yield_error(500, 'Sorry, our API seems to have issues connecting to our provider(s).', 'This most likely isn\'t your fault. Please try again later.')
|
yield await errors.yield_error(500, 'Sorry, our API seems to have issues connecting to our provider(s).', 'This most likely isn\'t your fault. Please try again later.')
|
||||||
|
|
|
@ -215,10 +215,7 @@ async def demo():
|
||||||
else:
|
else:
|
||||||
raise ConnectionError('API Server is not running.')
|
raise ConnectionError('API Server is not running.')
|
||||||
|
|
||||||
for func in [
|
for func in [test_chat_non_stream_gpt4, test_chat_stream_gpt3]:
|
||||||
# test_chat_non_stream_gpt4,
|
|
||||||
test_chat_stream_gpt3
|
|
||||||
]:
|
|
||||||
print(f'[*] {func.__name__}')
|
print(f'[*] {func.__name__}')
|
||||||
result = await func()
|
result = await func()
|
||||||
print(f'[+] {func.__name__} - {result}')
|
print(f'[+] {func.__name__} - {result}')
|
||||||
|
|
|
@ -1,10 +0,0 @@
|
||||||
--- EXPECTED ---
|
|
||||||
|
|
||||||
data: {"id":"custom-chatcmpl-nUSiapqELukaPT7vEnGcXkbvrS1fR","object":"chat.completion.chunk","created":1696716717,"model":"gpt-3.5-turbo-0613","choices":[{"index":0,"delta":{"role":"assistant","content":""},"finish_reason":null}]}
|
|
||||||
|
|
||||||
data: {"id":"custom-chatcmpl-nUSiapqELukaPT7vEnGcXkbvrS1fR","object":"chat.completion.chunk","created":1696716717,"model":"gpt-3.5-turbo-0613","choices":[{"index":0,"delta":{"content":"123"},"finish_reason":null}]}
|
|
||||||
|
|
||||||
data: {"id":"custom-chatcmpl-nUSiapqELukaPT7vEnGcXkbvrS1fR","object":"chat.completion.chunk","created":1696716717,"model":"gpt-3.5-turbo-0613","choices":[{"index":0,"delta":{},"finish_reason":"stop"}]}
|
|
||||||
|
|
||||||
data: [DONE]
|
|
||||||
|
|
Loading…
Reference in a new issue