Spaces:
Running
Running
Update main.py
Browse files
main.py
CHANGED
@@ -2,7 +2,7 @@ import os
|
|
2 |
import re
|
3 |
from dotenv import load_dotenv
|
4 |
from fastapi import FastAPI, HTTPException, Request, Depends, Security, Query
|
5 |
-
from fastapi.responses import StreamingResponse, HTMLResponse, JSONResponse, FileResponse,
|
6 |
from fastapi.security import APIKeyHeader
|
7 |
from pydantic import BaseModel
|
8 |
import httpx
|
@@ -22,25 +22,20 @@ from fastapi.middleware.gzip import GZipMiddleware
|
|
22 |
from starlette.middleware.cors import CORSMiddleware
|
23 |
import contextlib
|
24 |
import requests
|
25 |
-
|
26 |
asyncio.set_event_loop_policy(uvloop.EventLoopPolicy())
|
27 |
|
28 |
-
|
29 |
-
executor = ThreadPoolExecutor(max_workers=16) # Increased thread count for better parallelism
|
30 |
|
31 |
-
# Load environment variables once at startup
|
32 |
load_dotenv()
|
33 |
|
34 |
-
# API key security scheme
|
35 |
api_key_header = APIKeyHeader(name="Authorization", auto_error=False)
|
36 |
|
37 |
-
# Initialize usage tracker
|
38 |
from usage_tracker import UsageTracker
|
39 |
usage_tracker = UsageTracker()
|
40 |
|
41 |
app = FastAPI()
|
42 |
|
43 |
-
# Add middleware for compression and CORS
|
44 |
app.add_middleware(GZipMiddleware, minimum_size=1000)
|
45 |
app.add_middleware(
|
46 |
CORSMiddleware,
|
@@ -50,7 +45,6 @@ app.add_middleware(
|
|
50 |
allow_headers=["*"],
|
51 |
)
|
52 |
|
53 |
-
# Environment variables (cached)
|
54 |
@lru_cache(maxsize=1)
|
55 |
def get_env_vars():
|
56 |
return {
|
@@ -59,13 +53,14 @@ def get_env_vars():
|
|
59 |
'secret_api_endpoint_2': os.getenv('SECRET_API_ENDPOINT_2'),
|
60 |
'secret_api_endpoint_3': os.getenv('SECRET_API_ENDPOINT_3'),
|
61 |
'secret_api_endpoint_4': "https://text.pollinations.ai/openai",
|
62 |
-
'secret_api_endpoint_5': os.getenv('SECRET_API_ENDPOINT_5'),
|
|
|
63 |
'mistral_api': "https://api.mistral.ai",
|
64 |
'mistral_key': os.getenv('MISTRAL_KEY'),
|
|
|
65 |
'endpoint_origin': os.getenv('ENDPOINT_ORIGIN')
|
66 |
}
|
67 |
|
68 |
-
# Configuration for models - use sets for faster lookups
|
69 |
mistral_models = {
|
70 |
"mistral-large-latest",
|
71 |
"pixtral-large-latest",
|
@@ -115,7 +110,7 @@ alternate_models = {
|
|
115 |
"o3"
|
116 |
}
|
117 |
|
118 |
-
claude_3_models = {
|
119 |
"claude-3-7-sonnet",
|
120 |
"claude-3-7-sonnet-thinking",
|
121 |
"claude 3.5 haiku",
|
@@ -128,7 +123,19 @@ claude_3_models = { # Models for the new endpoint
|
|
128 |
"grok 2"
|
129 |
}
|
130 |
|
131 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
132 |
supported_image_models = {
|
133 |
"Flux Pro Ultra",
|
134 |
"grok-2-aurora",
|
@@ -143,86 +150,70 @@ supported_image_models = {
|
|
143 |
"sdxl-lightning-4step"
|
144 |
}
|
145 |
|
146 |
-
|
147 |
-
# Request payload model
|
148 |
class Payload(BaseModel):
|
149 |
model: str
|
150 |
messages: list
|
151 |
stream: bool = False
|
152 |
|
153 |
-
|
154 |
-
# Image generation payload model
|
155 |
class ImageGenerationPayload(BaseModel):
|
156 |
model: str
|
157 |
prompt: str
|
158 |
size: int
|
159 |
number: int
|
160 |
|
161 |
-
|
162 |
-
|
163 |
-
# Server status global variable
|
164 |
server_status = True
|
165 |
available_model_ids: List[str] = []
|
166 |
|
167 |
-
# Create a reusable httpx client pool with connection pooling
|
168 |
@lru_cache(maxsize=1)
|
169 |
def get_async_client():
|
170 |
return httpx.AsyncClient(
|
171 |
timeout=60.0,
|
172 |
-
limits=httpx.Limits(max_keepalive_connections=50, max_connections=200)
|
173 |
)
|
174 |
|
175 |
-
# Create a cloudscraper pool
|
176 |
scraper_pool = []
|
177 |
-
MAX_SCRAPERS = 20
|
178 |
-
|
179 |
|
180 |
def get_scraper():
|
181 |
if not scraper_pool:
|
182 |
for _ in range(MAX_SCRAPERS):
|
183 |
scraper_pool.append(cloudscraper.create_scraper())
|
184 |
|
185 |
-
return scraper_pool[int(time.time() * 1000) % MAX_SCRAPERS]
|
186 |
|
187 |
-
# API key validation - optimized to avoid string operations when possible
|
188 |
async def verify_api_key(
|
189 |
request: Request,
|
190 |
api_key: str = Security(api_key_header)
|
191 |
) -> bool:
|
192 |
-
# Allow bypass if the referer is from /playground or /image-playground
|
193 |
referer = request.headers.get("referer", "")
|
194 |
-
if referer.startswith(("https://parthsadaria-lokiai.hf.space/playground",
|
195 |
"https://parthsadaria-lokiai.hf.space/image-playground")):
|
196 |
return True
|
197 |
-
|
198 |
if not api_key:
|
199 |
raise HTTPException(
|
200 |
status_code=HTTP_403_FORBIDDEN,
|
201 |
detail="No API key provided"
|
202 |
)
|
203 |
-
|
204 |
-
# Only clean if needed
|
205 |
if api_key.startswith('Bearer '):
|
206 |
-
api_key = api_key[7:]
|
207 |
-
|
208 |
-
# Get API keys from environment
|
209 |
valid_api_keys = get_env_vars().get('api_keys', [])
|
210 |
if not valid_api_keys or valid_api_keys == ['']:
|
211 |
raise HTTPException(
|
212 |
status_code=HTTP_403_FORBIDDEN,
|
213 |
detail="API keys not configured on server"
|
214 |
)
|
215 |
-
|
216 |
-
# Fast check with set operation
|
217 |
if api_key not in set(valid_api_keys):
|
218 |
raise HTTPException(
|
219 |
status_code=HTTP_403_FORBIDDEN,
|
220 |
detail="Invalid API key"
|
221 |
)
|
222 |
-
|
223 |
return True
|
224 |
|
225 |
-
# Pre-load and cache models.json
|
226 |
@lru_cache(maxsize=1)
|
227 |
def load_models_data():
|
228 |
try:
|
@@ -233,61 +224,44 @@ def load_models_data():
|
|
233 |
print(f"Error loading models.json: {str(e)}")
|
234 |
return []
|
235 |
|
236 |
-
# Async wrapper for models data
|
237 |
async def get_models():
|
238 |
models_data = load_models_data()
|
239 |
if not models_data:
|
240 |
raise HTTPException(status_code=500, detail="Error loading available models")
|
241 |
return models_data
|
242 |
|
243 |
-
# Enhanced async streaming - now with real-time SSE support
|
244 |
async def generate_search_async(query: str, systemprompt: Optional[str] = None, stream: bool = True):
|
245 |
-
# Create a streaming response channel using asyncio.Queue
|
246 |
queue = asyncio.Queue()
|
247 |
|
248 |
async def _fetch_search_data():
|
249 |
try:
|
250 |
headers = {"User-Agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36"}
|
251 |
-
|
252 |
-
# Use the provided system prompt, or default to "Be Helpful and Friendly"
|
253 |
system_message = systemprompt or "Be Helpful and Friendly"
|
254 |
-
|
255 |
-
# Create the prompt history
|
256 |
-
prompt = [
|
257 |
-
{"role": "user", "content": query},
|
258 |
-
]
|
259 |
-
|
260 |
prompt.insert(0, {"content": system_message, "role": "system"})
|
261 |
-
|
262 |
-
# Prepare the payload for the API request
|
263 |
payload = {
|
264 |
"is_vscode_extension": True,
|
265 |
"message_history": prompt,
|
266 |
"requested_model": "searchgpt",
|
267 |
"user_input": prompt[-1]["content"],
|
268 |
}
|
269 |
-
|
270 |
-
# Get endpoint from environment
|
271 |
secret_api_endpoint_3 = get_env_vars()['secret_api_endpoint_3']
|
272 |
if not secret_api_endpoint_3:
|
273 |
await queue.put({"error": "Search API endpoint not configured"})
|
274 |
return
|
275 |
|
276 |
-
# Use AsyncClient for better performance
|
277 |
async with httpx.AsyncClient(timeout=30.0) as client:
|
278 |
async with client.stream("POST", secret_api_endpoint_3, json=payload, headers=headers) as response:
|
279 |
if response.status_code != 200:
|
280 |
await queue.put({"error": f"Search API returned status code {response.status_code}"})
|
281 |
return
|
282 |
|
283 |
-
# Process the streaming response in real-time
|
284 |
buffer = ""
|
285 |
async for line in response.aiter_lines():
|
286 |
if line.startswith("data: "):
|
287 |
try:
|
288 |
json_data = json.loads(line[6:])
|
289 |
content = json_data.get("choices", [{}])[0].get("delta", {}).get("content", "")
|
290 |
-
|
291 |
if content.strip():
|
292 |
cleaned_response = {
|
293 |
"created": json_data.get("created"),
|
@@ -302,26 +276,17 @@ async def generate_search_async(query: str, systemprompt: Optional[str] = None,
|
|
302 |
}
|
303 |
]
|
304 |
}
|
305 |
-
|
306 |
-
# Send to queue immediately for streaming
|
307 |
await queue.put({"data": f"data: {json.dumps(cleaned_response)}\n\n", "text": content})
|
308 |
except json.JSONDecodeError:
|
309 |
continue
|
310 |
-
|
311 |
-
# Signal completion
|
312 |
await queue.put(None)
|
313 |
-
|
314 |
except Exception as e:
|
315 |
await queue.put({"error": str(e)})
|
316 |
await queue.put(None)
|
317 |
|
318 |
-
# Start the fetch process
|
319 |
asyncio.create_task(_fetch_search_data())
|
320 |
-
|
321 |
-
# Return the queue for consumption
|
322 |
return queue
|
323 |
|
324 |
-
# Cache for frequently accessed static files
|
325 |
@lru_cache(maxsize=10)
|
326 |
def read_html_file(file_path):
|
327 |
try:
|
@@ -330,16 +295,15 @@ def read_html_file(file_path):
|
|
330 |
except FileNotFoundError:
|
331 |
return None
|
332 |
|
333 |
-
# Basic routes
|
334 |
@app.get("/favicon.ico")
|
335 |
async def favicon():
|
336 |
favicon_path = Path(__file__).parent / "favicon.ico"
|
337 |
return FileResponse(favicon_path, media_type="image/x-icon")
|
338 |
|
339 |
@app.get("/banner.jpg")
|
340 |
-
async def
|
341 |
-
|
342 |
-
return FileResponse(
|
343 |
|
344 |
@app.get("/ping")
|
345 |
async def ping():
|
@@ -351,64 +315,61 @@ async def root():
|
|
351 |
if html_content is None:
|
352 |
return HTMLResponse(content="<h1>File not found</h1>", status_code=404)
|
353 |
return HTMLResponse(content=html_content)
|
|
|
354 |
@app.get("/script.js", response_class=HTMLResponse)
|
355 |
-
async def
|
356 |
html_content = read_html_file("script.js")
|
357 |
if html_content is None:
|
358 |
return HTMLResponse(content="<h1>File not found</h1>", status_code=404)
|
359 |
return HTMLResponse(content=html_content)
|
|
|
360 |
@app.get("/style.css", response_class=HTMLResponse)
|
361 |
-
async def
|
362 |
html_content = read_html_file("style.css")
|
363 |
if html_content is None:
|
364 |
return HTMLResponse(content="<h1>File not found</h1>", status_code=404)
|
365 |
return HTMLResponse(content=html_content)
|
|
|
366 |
@app.get("/dynamo", response_class=HTMLResponse)
|
367 |
async def dynamic_ai_page(request: Request):
|
368 |
user_agent = request.headers.get('user-agent', 'Unknown User')
|
369 |
client_ip = request.client.host
|
370 |
location = f"IP: {client_ip}"
|
371 |
-
|
372 |
prompt = f"""
|
373 |
-
Generate a dynamic HTML page for a user with the following details: with name "LOKI.AI"
|
374 |
- User-Agent: {user_agent}
|
375 |
- Location: {location}
|
376 |
- Style: Cyberpunk, minimalist, or retro
|
377 |
-
|
378 |
Make sure the HTML is clean and includes a heading, also have cool animations a motivational message, and a cool background.
|
379 |
Wrap the generated HTML in triple backticks (```).
|
380 |
"""
|
381 |
-
|
382 |
payload = {
|
383 |
"model": "mistral-small-latest",
|
384 |
"messages": [{"role": "user", "content": prompt}]
|
385 |
}
|
386 |
-
|
387 |
headers = {
|
388 |
"Authorization": "Bearer playground"
|
389 |
}
|
390 |
-
|
391 |
-
response = requests.post("https://parthsadaria-lokiai.hf.space/chat/completions", json=payload, headers=headers)
|
392 |
data = response.json()
|
393 |
-
|
394 |
-
# Extract HTML from ``` blocks
|
395 |
html_content = re.search(r"```(.*?)```", data['choices'][0]['message']['content'], re.DOTALL)
|
396 |
if html_content:
|
397 |
html_content = html_content.group(1).strip()
|
398 |
-
|
399 |
-
# Remove the first word
|
400 |
if html_content:
|
401 |
html_content = ' '.join(html_content.split(' ')[1:])
|
402 |
-
|
403 |
-
return HTMLResponse(content=html_content)
|
404 |
|
405 |
-
|
406 |
-
######################################
|
407 |
|
408 |
@app.get("/scraper", response_class=PlainTextResponse)
|
409 |
def scrape_site(url: str = Query(..., description="URL to scrape")):
|
410 |
try:
|
411 |
-
# Try cloudscraper first
|
412 |
scraper = cloudscraper.create_scraper()
|
413 |
response = scraper.get(url)
|
414 |
if response.status_code == 200 and len(response.text.strip()) > 0:
|
@@ -417,28 +378,21 @@ def scrape_site(url: str = Query(..., description="URL to scrape")):
|
|
417 |
print(f"Cloudscraper failed: {e}")
|
418 |
return "Cloudscraper failed."
|
419 |
|
420 |
-
|
421 |
-
#######################################
|
422 |
-
|
423 |
@app.get("/playground", response_class=HTMLResponse)
|
424 |
async def playground():
|
425 |
html_content = read_html_file("playground.html")
|
426 |
if html_content is None:
|
427 |
return HTMLResponse(content="<h1>playground.html not found</h1>", status_code=404)
|
428 |
return HTMLResponse(content=html_content)
|
429 |
-
|
430 |
@app.get("/image-playground", response_class=HTMLResponse)
|
431 |
-
async def
|
432 |
html_content = read_html_file("image-playground.html")
|
433 |
if html_content is None:
|
434 |
return HTMLResponse(content="<h1>image-playground.html not found</h1>", status_code=404)
|
435 |
return HTMLResponse(content=html_content)
|
436 |
|
437 |
-
|
438 |
-
|
439 |
-
|
440 |
-
# VETRA
|
441 |
-
GITHUB_BASE = "https://raw.githubusercontent.com/Parthsadaria/Vetra/main"
|
442 |
|
443 |
FILES = {
|
444 |
"html": "index.html",
|
@@ -471,17 +425,11 @@ async def serve_vetra():
|
|
471 |
|
472 |
return HTMLResponse(content=final_html)
|
473 |
|
474 |
-
|
475 |
-
|
476 |
-
|
477 |
-
|
478 |
-
# Model routes
|
479 |
@app.get("/api/v1/models")
|
480 |
@app.get("/models")
|
481 |
async def return_models():
|
482 |
return await get_models()
|
483 |
|
484 |
-
# Search routes with enhanced real-time streaming
|
485 |
@app.get("/searchgpt")
|
486 |
async def search_gpt(q: str, stream: Optional[bool] = False, systemprompt: Optional[str] = None):
|
487 |
if not q:
|
@@ -512,7 +460,6 @@ async def search_gpt(q: str, stream: Optional[bool] = False, systemprompt: Optio
|
|
512 |
media_type="text/event-stream"
|
513 |
)
|
514 |
else:
|
515 |
-
# For non-streaming, collect all text and return at once
|
516 |
collected_text = ""
|
517 |
while True:
|
518 |
item = await queue.get()
|
@@ -526,14 +473,10 @@ async def search_gpt(q: str, stream: Optional[bool] = False, systemprompt: Optio
|
|
526 |
|
527 |
return JSONResponse(content={"response": collected_text})
|
528 |
|
529 |
-
|
530 |
-
|
531 |
-
# Enhanced streaming with direct SSE pass-through for real-time responses
|
532 |
header_url = os.getenv('HEADER_URL')
|
533 |
@app.post("/chat/completions")
|
534 |
@app.post("/api/v1/chat/completions")
|
535 |
async def get_completion(payload: Payload, request: Request, authenticated: bool = Depends(verify_api_key)):
|
536 |
-
# Check server status
|
537 |
if not server_status:
|
538 |
return JSONResponse(
|
539 |
status_code=503,
|
@@ -542,28 +485,22 @@ async def get_completion(payload: Payload, request: Request, authenticated: bool
|
|
542 |
|
543 |
model_to_use = payload.model or "gpt-4o-mini"
|
544 |
|
545 |
-
# Validate model availability - fast lookup with set
|
546 |
if available_model_ids and model_to_use not in set(available_model_ids):
|
547 |
raise HTTPException(
|
548 |
status_code=400,
|
549 |
detail=f"Model '{model_to_use}' is not available. Check /models for the available model list."
|
550 |
)
|
551 |
|
552 |
-
# Log request without blocking
|
553 |
asyncio.create_task(log_request(request, model_to_use))
|
554 |
usage_tracker.record_request(model=model_to_use, endpoint="/chat/completions")
|
555 |
|
556 |
-
# Prepare payload
|
557 |
payload_dict = payload.dict()
|
558 |
payload_dict["model"] = model_to_use
|
559 |
|
560 |
-
# Ensure stream is True for real-time streaming (can be overridden by client)
|
561 |
stream_enabled = payload_dict.get("stream", True)
|
562 |
|
563 |
-
# Get environment variables
|
564 |
env_vars = get_env_vars()
|
565 |
|
566 |
-
# Select the appropriate endpoint (fast lookup with sets)
|
567 |
if model_to_use in mistral_models:
|
568 |
endpoint = env_vars['mistral_api']
|
569 |
custom_headers = {
|
@@ -575,9 +512,18 @@ async def get_completion(payload: Payload, request: Request, authenticated: bool
|
|
575 |
elif model_to_use in alternate_models:
|
576 |
endpoint = env_vars['secret_api_endpoint_2']
|
577 |
custom_headers = {}
|
578 |
-
elif model_to_use in claude_3_models:
|
579 |
endpoint = env_vars['secret_api_endpoint_5']
|
580 |
custom_headers = {}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
581 |
else:
|
582 |
endpoint = env_vars['secret_api_endpoint']
|
583 |
custom_headers = {
|
@@ -588,7 +534,6 @@ async def get_completion(payload: Payload, request: Request, authenticated: bool
|
|
588 |
|
589 |
print(f"Using endpoint: {endpoint} for model: {model_to_use}")
|
590 |
|
591 |
-
# Improved real-time streaming handler
|
592 |
async def real_time_stream_generator():
|
593 |
try:
|
594 |
async with httpx.AsyncClient(timeout=60.0) as client:
|
@@ -603,10 +548,8 @@ async def get_completion(payload: Payload, request: Request, authenticated: bool
|
|
603 |
detail = error_messages.get(response.status_code, f"Error code: {response.status_code}")
|
604 |
raise HTTPException(status_code=response.status_code, detail=detail)
|
605 |
|
606 |
-
# Stream the response in real-time with minimal buffering
|
607 |
async for line in response.aiter_lines():
|
608 |
if line:
|
609 |
-
# Yield immediately for faster streaming
|
610 |
yield line + "\n"
|
611 |
except httpx.TimeoutException:
|
612 |
raise HTTPException(status_code=504, detail="Request timed out")
|
@@ -617,7 +560,6 @@ async def get_completion(payload: Payload, request: Request, authenticated: bool
|
|
617 |
raise e
|
618 |
raise HTTPException(status_code=500, detail=f"An error occurred: {str(e)}")
|
619 |
|
620 |
-
# Return streaming response with proper headers
|
621 |
if stream_enabled:
|
622 |
return StreamingResponse(
|
623 |
real_time_stream_generator(),
|
@@ -626,43 +568,31 @@ async def get_completion(payload: Payload, request: Request, authenticated: bool
|
|
626 |
"Content-Type": "text/event-stream",
|
627 |
"Cache-Control": "no-cache",
|
628 |
"Connection": "keep-alive",
|
629 |
-
"X-Accel-Buffering": "no"
|
630 |
}
|
631 |
)
|
632 |
else:
|
633 |
-
# For non-streaming requests, collect the entire response
|
634 |
response_content = []
|
635 |
async for chunk in real_time_stream_generator():
|
636 |
response_content.append(chunk)
|
637 |
-
|
638 |
return JSONResponse(content=json.loads(''.join(response_content)))
|
639 |
|
640 |
-
|
641 |
-
|
642 |
-
# New image generation endpoint
|
643 |
@app.post("/images/generations")
|
644 |
async def create_image(payload: ImageGenerationPayload, authenticated: bool = Depends(verify_api_key)):
|
645 |
-
"""
|
646 |
-
Endpoint for generating images based on a text prompt.
|
647 |
-
"""
|
648 |
-
# Check server status
|
649 |
if not server_status:
|
650 |
return JSONResponse(
|
651 |
status_code=503,
|
652 |
content={"message": "Server is under maintenance. Please try again later."}
|
653 |
)
|
654 |
|
655 |
-
# Validate model
|
656 |
if payload.model not in supported_image_models:
|
657 |
raise HTTPException(
|
658 |
status_code=400,
|
659 |
-
detail=f"Model '{payload.model}' is not supported for image generation.
|
660 |
)
|
661 |
|
662 |
-
# Log the request
|
663 |
usage_tracker.record_request(model=payload.model, endpoint="/images/generations")
|
664 |
|
665 |
-
# Prepare the payload for the external API
|
666 |
api_payload = {
|
667 |
"model": payload.model,
|
668 |
"prompt": payload.prompt,
|
@@ -670,11 +600,9 @@ async def create_image(payload: ImageGenerationPayload, authenticated: bool = De
|
|
670 |
"number": payload.number
|
671 |
}
|
672 |
|
673 |
-
# Target API endpoint
|
674 |
target_api_url = os.getenv('NEW_IMG')
|
675 |
|
676 |
try:
|
677 |
-
# Use a timeout for the image generation request
|
678 |
async with httpx.AsyncClient(timeout=60.0) as client:
|
679 |
response = await client.post(target_api_url, json=api_payload)
|
680 |
|
@@ -682,7 +610,6 @@ async def create_image(payload: ImageGenerationPayload, authenticated: bool = De
|
|
682 |
error_detail = response.json().get("detail", f"Image generation failed with status code: {response.status_code}")
|
683 |
raise HTTPException(status_code=response.status_code, detail=error_detail)
|
684 |
|
685 |
-
# Return the response from the external API
|
686 |
return JSONResponse(content=response.json())
|
687 |
|
688 |
except httpx.TimeoutException:
|
@@ -692,28 +619,20 @@ async def create_image(payload: ImageGenerationPayload, authenticated: bool = De
|
|
692 |
except Exception as e:
|
693 |
raise HTTPException(status_code=500, detail=f"An unexpected error occurred during image generation: {e}")
|
694 |
|
695 |
-
|
696 |
-
|
697 |
-
# Asynchronous logging function
|
698 |
async def log_request(request, model):
|
699 |
-
# Get minimal data for logging
|
700 |
current_time = (datetime.datetime.utcnow() + datetime.timedelta(hours=5, minutes=30)).strftime("%Y-%m-%d %I:%M:%S %p")
|
701 |
-
ip_hash = hash(request.client.host) % 10000
|
702 |
print(f"Time: {current_time}, IP Hash: {ip_hash}, Model: {model}")
|
703 |
|
704 |
-
# Cache usage statistics
|
705 |
@lru_cache(maxsize=10)
|
706 |
def get_usage_summary(days=7):
|
707 |
return usage_tracker.get_usage_summary(days)
|
708 |
|
709 |
@app.get("/usage")
|
710 |
async def get_usage(days: int = 7):
|
711 |
-
"""Retrieve usage statistics"""
|
712 |
return get_usage_summary(days)
|
713 |
|
714 |
-
# Generate HTML for usage page
|
715 |
def generate_usage_html(usage_data):
|
716 |
-
# Model Usage Table Rows
|
717 |
model_usage_rows = "\n".join([
|
718 |
f"""
|
719 |
<tr>
|
@@ -725,7 +644,6 @@ def generate_usage_html(usage_data):
|
|
725 |
""" for model, model_data in usage_data['models'].items()
|
726 |
])
|
727 |
|
728 |
-
# API Endpoint Usage Table Rows
|
729 |
api_usage_rows = "\n".join([
|
730 |
f"""
|
731 |
<tr>
|
@@ -737,7 +655,6 @@ def generate_usage_html(usage_data):
|
|
737 |
""" for endpoint, endpoint_data in usage_data['api_endpoints'].items()
|
738 |
])
|
739 |
|
740 |
-
# Daily Usage Table Rows
|
741 |
daily_usage_rows = "\n".join([
|
742 |
"\n".join([
|
743 |
f"""
|
@@ -756,7 +673,7 @@ def generate_usage_html(usage_data):
|
|
756 |
<head>
|
757 |
<meta charset="UTF-8">
|
758 |
<title>Lokiai AI - Usage Statistics</title>
|
759 |
-
<link href="https://fonts.googleapis.com/css2?family=Inter:wght@300;400;600&display=swap" rel="stylesheet">
|
760 |
<style>
|
761 |
:root {{
|
762 |
--bg-dark: #0f1011;
|
@@ -902,7 +819,6 @@ def generate_usage_html(usage_data):
|
|
902 |
"""
|
903 |
return html_content
|
904 |
|
905 |
-
# Cache the usage page HTML
|
906 |
@lru_cache(maxsize=1)
|
907 |
def get_usage_page_html():
|
908 |
usage_data = get_usage_summary()
|
@@ -910,18 +826,14 @@ def get_usage_page_html():
|
|
910 |
|
911 |
@app.get("/usage/page", response_class=HTMLResponse)
|
912 |
async def usage_page():
|
913 |
-
"""Serve an HTML page showing usage statistics"""
|
914 |
-
# Use cached HTML if available, regenerate if not
|
915 |
html_content = get_usage_page_html()
|
916 |
return HTMLResponse(content=html_content)
|
917 |
|
918 |
-
# Meme endpoint with optimized networking
|
919 |
@app.get("/meme")
|
920 |
async def get_meme():
|
921 |
try:
|
922 |
-
# Use the shared client for connection pooling
|
923 |
client = get_async_client()
|
924 |
-
response = await client.get("https://meme-api.com/gimme")
|
925 |
response_data = response.json()
|
926 |
|
927 |
meme_url = response_data.get("url")
|
@@ -930,36 +842,31 @@ async def get_meme():
|
|
930 |
|
931 |
image_response = await client.get(meme_url, follow_redirects=True)
|
932 |
|
933 |
-
# Use larger chunks for streaming
|
934 |
async def stream_with_larger_chunks():
|
935 |
chunks = []
|
936 |
size = 0
|
937 |
async for chunk in image_response.aiter_bytes(chunk_size=16384):
|
938 |
chunks.append(chunk)
|
939 |
size += len(chunk)
|
940 |
-
|
941 |
if size >= 65536:
|
942 |
yield b''.join(chunks)
|
943 |
chunks = []
|
944 |
size = 0
|
945 |
-
|
946 |
if chunks:
|
947 |
yield b''.join(chunks)
|
948 |
|
949 |
return StreamingResponse(
|
950 |
stream_with_larger_chunks(),
|
951 |
media_type=image_response.headers.get("content-type", "image/png"),
|
952 |
-
headers={'Cache-Control': 'max-age=3600'}
|
953 |
)
|
954 |
except Exception:
|
955 |
raise HTTPException(status_code=500, detail="Failed to retrieve meme")
|
956 |
|
957 |
-
# Utility function for loading model IDs - optimized to run once at startup
|
958 |
def load_model_ids(json_file_path):
|
959 |
try:
|
960 |
with open(json_file_path, 'r') as f:
|
961 |
models_data = json.load(f)
|
962 |
-
# Extract 'id' from each model object and use a set for fast lookups
|
963 |
return [model['id'] for model in models_data if 'id' in model]
|
964 |
except Exception as e:
|
965 |
print(f"Error loading model IDs: {str(e)}")
|
@@ -971,23 +878,18 @@ async def startup_event():
|
|
971 |
available_model_ids = load_model_ids("models.json")
|
972 |
print(f"Loaded {len(available_model_ids)} model IDs")
|
973 |
|
974 |
-
# Add all pollinations models to available_model_ids
|
975 |
available_model_ids.extend(list(pollinations_models))
|
976 |
-
# Add alternate models to available_model_ids
|
977 |
available_model_ids.extend(list(alternate_models))
|
978 |
-
# Add mistral models to available_model_ids
|
979 |
available_model_ids.extend(list(mistral_models))
|
980 |
-
# Add claude models
|
981 |
available_model_ids.extend(list(claude_3_models))
|
|
|
982 |
|
983 |
-
available_model_ids = list(set(available_model_ids))
|
984 |
print(f"Total available models: {len(available_model_ids)}")
|
985 |
|
986 |
-
# Preload scrapers
|
987 |
for _ in range(MAX_SCRAPERS):
|
988 |
scraper_pool.append(cloudscraper.create_scraper())
|
989 |
|
990 |
-
# Validate critical environment variables
|
991 |
env_vars = get_env_vars()
|
992 |
missing_vars = []
|
993 |
|
@@ -1001,12 +903,16 @@ async def startup_event():
|
|
1001 |
missing_vars.append('SECRET_API_ENDPOINT_3')
|
1002 |
if not env_vars['secret_api_endpoint_4']:
|
1003 |
missing_vars.append('SECRET_API_ENDPOINT_4')
|
1004 |
-
if not env_vars['secret_api_endpoint_5']:
|
1005 |
missing_vars.append('SECRET_API_ENDPOINT_5')
|
|
|
|
|
1006 |
if not env_vars['mistral_api'] and any(model in mistral_models for model in available_model_ids):
|
1007 |
missing_vars.append('MISTRAL_API')
|
1008 |
if not env_vars['mistral_key'] and any(model in mistral_models for model in available_model_ids):
|
1009 |
missing_vars.append('MISTRAL_KEY')
|
|
|
|
|
1010 |
|
1011 |
if missing_vars:
|
1012 |
print(f"WARNING: The following environment variables are missing: {', '.join(missing_vars)}")
|
@@ -1016,27 +922,17 @@ async def startup_event():
|
|
1016 |
|
1017 |
@app.on_event("shutdown")
|
1018 |
async def shutdown_event():
|
1019 |
-
# Close the httpx client
|
1020 |
client = get_async_client()
|
1021 |
await client.aclose()
|
1022 |
-
|
1023 |
-
# Clear scraper pool
|
1024 |
scraper_pool.clear()
|
1025 |
-
|
1026 |
-
# Persist usage data
|
1027 |
usage_tracker.save_data()
|
1028 |
-
|
1029 |
print("Server shutdown complete!")
|
1030 |
|
1031 |
-
# Health check endpoint
|
1032 |
-
# Health check endpoint
|
1033 |
@app.get("/health")
|
1034 |
async def health_check():
|
1035 |
-
"""Health check endpoint for monitoring"""
|
1036 |
env_vars = get_env_vars()
|
1037 |
missing_critical_vars = []
|
1038 |
|
1039 |
-
# Check critical environment variables
|
1040 |
if not env_vars['api_keys'] or env_vars['api_keys'] == ['']:
|
1041 |
missing_critical_vars.append('API_KEYS')
|
1042 |
if not env_vars['secret_api_endpoint']:
|
@@ -1047,12 +943,16 @@ async def health_check():
|
|
1047 |
missing_critical_vars.append('SECRET_API_ENDPOINT_3')
|
1048 |
if not env_vars['secret_api_endpoint_4']:
|
1049 |
missing_critical_vars.append('SECRET_API_ENDPOINT_4')
|
1050 |
-
if not env_vars['secret_api_endpoint_5']:
|
1051 |
missing_critical_vars.append('SECRET_API_ENDPOINT_5')
|
|
|
|
|
1052 |
if not env_vars['mistral_api']:
|
1053 |
missing_critical_vars.append('MISTRAL_API')
|
1054 |
if not env_vars['mistral_key']:
|
1055 |
missing_critical_vars.append('MISTRAL_KEY')
|
|
|
|
|
1056 |
|
1057 |
health_status = {
|
1058 |
"status": "healthy" if not missing_critical_vars else "unhealthy",
|
@@ -1064,4 +964,4 @@ async def health_check():
|
|
1064 |
|
1065 |
if __name__ == "__main__":
|
1066 |
import uvicorn
|
1067 |
-
uvicorn.run(app, host="0.0.0.0", port=7860)
|
|
|
2 |
import re
|
3 |
from dotenv import load_dotenv
|
4 |
from fastapi import FastAPI, HTTPException, Request, Depends, Security, Query
|
5 |
+
from fastapi.responses import StreamingResponse, HTMLResponse, JSONResponse, FileResponse, PlainTextResponse
|
6 |
from fastapi.security import APIKeyHeader
|
7 |
from pydantic import BaseModel
|
8 |
import httpx
|
|
|
22 |
from starlette.middleware.cors import CORSMiddleware
|
23 |
import contextlib
|
24 |
import requests
|
25 |
+
|
26 |
asyncio.set_event_loop_policy(uvloop.EventLoopPolicy())
|
27 |
|
28 |
+
executor = ThreadPoolExecutor(max_workers=16)
|
|
|
29 |
|
|
|
30 |
load_dotenv()
|
31 |
|
|
|
32 |
api_key_header = APIKeyHeader(name="Authorization", auto_error=False)
|
33 |
|
|
|
34 |
from usage_tracker import UsageTracker
|
35 |
usage_tracker = UsageTracker()
|
36 |
|
37 |
app = FastAPI()
|
38 |
|
|
|
39 |
app.add_middleware(GZipMiddleware, minimum_size=1000)
|
40 |
app.add_middleware(
|
41 |
CORSMiddleware,
|
|
|
45 |
allow_headers=["*"],
|
46 |
)
|
47 |
|
|
|
48 |
@lru_cache(maxsize=1)
|
49 |
def get_env_vars():
|
50 |
return {
|
|
|
53 |
'secret_api_endpoint_2': os.getenv('SECRET_API_ENDPOINT_2'),
|
54 |
'secret_api_endpoint_3': os.getenv('SECRET_API_ENDPOINT_3'),
|
55 |
'secret_api_endpoint_4': "https://text.pollinations.ai/openai",
|
56 |
+
'secret_api_endpoint_5': os.getenv('SECRET_API_ENDPOINT_5'),
|
57 |
+
'secret_api_endpoint_6': os.getenv('SECRET_API_ENDPOINT_6'), # New endpoint for Gemini
|
58 |
'mistral_api': "https://api.mistral.ai",
|
59 |
'mistral_key': os.getenv('MISTRAL_KEY'),
|
60 |
+
'gemini_key': os.getenv('GEMINI_KEY'), # Gemini API Key
|
61 |
'endpoint_origin': os.getenv('ENDPOINT_ORIGIN')
|
62 |
}
|
63 |
|
|
|
64 |
mistral_models = {
|
65 |
"mistral-large-latest",
|
66 |
"pixtral-large-latest",
|
|
|
110 |
"o3"
|
111 |
}
|
112 |
|
113 |
+
claude_3_models = {
|
114 |
"claude-3-7-sonnet",
|
115 |
"claude-3-7-sonnet-thinking",
|
116 |
"claude 3.5 haiku",
|
|
|
123 |
"grok 2"
|
124 |
}
|
125 |
|
126 |
+
gemini_models = {
|
127 |
+
"gemini-1.5-pro",
|
128 |
+
"gemini-1.5-flash",
|
129 |
+
"gemini-2.0-flash-lite-preview",
|
130 |
+
"gemini-2.0-flash",
|
131 |
+
"gemini-2.0-flash-thinking", # aka Reasoning
|
132 |
+
"gemini-2.0-flash-preview-image-generation",
|
133 |
+
"gemini-2.5-flash",
|
134 |
+
"gemini-2.5-pro-exp",
|
135 |
+
"gemini-exp-1206"
|
136 |
+
}
|
137 |
+
|
138 |
+
|
139 |
supported_image_models = {
|
140 |
"Flux Pro Ultra",
|
141 |
"grok-2-aurora",
|
|
|
150 |
"sdxl-lightning-4step"
|
151 |
}
|
152 |
|
|
|
|
|
153 |
class Payload(BaseModel):
|
154 |
model: str
|
155 |
messages: list
|
156 |
stream: bool = False
|
157 |
|
|
|
|
|
158 |
class ImageGenerationPayload(BaseModel):
|
159 |
model: str
|
160 |
prompt: str
|
161 |
size: int
|
162 |
number: int
|
163 |
|
|
|
|
|
|
|
164 |
server_status = True
|
165 |
available_model_ids: List[str] = []
|
166 |
|
|
|
167 |
@lru_cache(maxsize=1)
|
168 |
def get_async_client():
|
169 |
return httpx.AsyncClient(
|
170 |
timeout=60.0,
|
171 |
+
limits=httpx.Limits(max_keepalive_connections=50, max_connections=200)
|
172 |
)
|
173 |
|
|
|
174 |
scraper_pool = []
|
175 |
+
MAX_SCRAPERS = 20
|
|
|
176 |
|
177 |
def get_scraper():
|
178 |
if not scraper_pool:
|
179 |
for _ in range(MAX_SCRAPERS):
|
180 |
scraper_pool.append(cloudscraper.create_scraper())
|
181 |
|
182 |
+
return scraper_pool[int(time.time() * 1000) % MAX_SCRAPERS]
|
183 |
|
|
|
184 |
async def verify_api_key(
|
185 |
request: Request,
|
186 |
api_key: str = Security(api_key_header)
|
187 |
) -> bool:
|
|
|
188 |
referer = request.headers.get("referer", "")
|
189 |
+
if referer.startswith(("https://parthsadaria-lokiai.hf.space/playground",
|
190 |
"https://parthsadaria-lokiai.hf.space/image-playground")):
|
191 |
return True
|
192 |
+
|
193 |
if not api_key:
|
194 |
raise HTTPException(
|
195 |
status_code=HTTP_403_FORBIDDEN,
|
196 |
detail="No API key provided"
|
197 |
)
|
198 |
+
|
|
|
199 |
if api_key.startswith('Bearer '):
|
200 |
+
api_key = api_key[7:]
|
201 |
+
|
|
|
202 |
valid_api_keys = get_env_vars().get('api_keys', [])
|
203 |
if not valid_api_keys or valid_api_keys == ['']:
|
204 |
raise HTTPException(
|
205 |
status_code=HTTP_403_FORBIDDEN,
|
206 |
detail="API keys not configured on server"
|
207 |
)
|
208 |
+
|
|
|
209 |
if api_key not in set(valid_api_keys):
|
210 |
raise HTTPException(
|
211 |
status_code=HTTP_403_FORBIDDEN,
|
212 |
detail="Invalid API key"
|
213 |
)
|
214 |
+
|
215 |
return True
|
216 |
|
|
|
217 |
@lru_cache(maxsize=1)
|
218 |
def load_models_data():
|
219 |
try:
|
|
|
224 |
print(f"Error loading models.json: {str(e)}")
|
225 |
return []
|
226 |
|
|
|
227 |
async def get_models():
|
228 |
models_data = load_models_data()
|
229 |
if not models_data:
|
230 |
raise HTTPException(status_code=500, detail="Error loading available models")
|
231 |
return models_data
|
232 |
|
|
|
233 |
async def generate_search_async(query: str, systemprompt: Optional[str] = None, stream: bool = True):
|
|
|
234 |
queue = asyncio.Queue()
|
235 |
|
236 |
async def _fetch_search_data():
|
237 |
try:
|
238 |
headers = {"User-Agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36"}
|
|
|
|
|
239 |
system_message = systemprompt or "Be Helpful and Friendly"
|
240 |
+
prompt = [{"role": "user", "content": query}]
|
|
|
|
|
|
|
|
|
|
|
241 |
prompt.insert(0, {"content": system_message, "role": "system"})
|
|
|
|
|
242 |
payload = {
|
243 |
"is_vscode_extension": True,
|
244 |
"message_history": prompt,
|
245 |
"requested_model": "searchgpt",
|
246 |
"user_input": prompt[-1]["content"],
|
247 |
}
|
|
|
|
|
248 |
secret_api_endpoint_3 = get_env_vars()['secret_api_endpoint_3']
|
249 |
if not secret_api_endpoint_3:
|
250 |
await queue.put({"error": "Search API endpoint not configured"})
|
251 |
return
|
252 |
|
|
|
253 |
async with httpx.AsyncClient(timeout=30.0) as client:
|
254 |
async with client.stream("POST", secret_api_endpoint_3, json=payload, headers=headers) as response:
|
255 |
if response.status_code != 200:
|
256 |
await queue.put({"error": f"Search API returned status code {response.status_code}"})
|
257 |
return
|
258 |
|
|
|
259 |
buffer = ""
|
260 |
async for line in response.aiter_lines():
|
261 |
if line.startswith("data: "):
|
262 |
try:
|
263 |
json_data = json.loads(line[6:])
|
264 |
content = json_data.get("choices", [{}])[0].get("delta", {}).get("content", "")
|
|
|
265 |
if content.strip():
|
266 |
cleaned_response = {
|
267 |
"created": json_data.get("created"),
|
|
|
276 |
}
|
277 |
]
|
278 |
}
|
|
|
|
|
279 |
await queue.put({"data": f"data: {json.dumps(cleaned_response)}\n\n", "text": content})
|
280 |
except json.JSONDecodeError:
|
281 |
continue
|
|
|
|
|
282 |
await queue.put(None)
|
|
|
283 |
except Exception as e:
|
284 |
await queue.put({"error": str(e)})
|
285 |
await queue.put(None)
|
286 |
|
|
|
287 |
asyncio.create_task(_fetch_search_data())
|
|
|
|
|
288 |
return queue
|
289 |
|
|
|
290 |
@lru_cache(maxsize=10)
|
291 |
def read_html_file(file_path):
|
292 |
try:
|
|
|
295 |
except FileNotFoundError:
|
296 |
return None
|
297 |
|
|
|
298 |
@app.get("/favicon.ico")
|
299 |
async def favicon():
|
300 |
favicon_path = Path(__file__).parent / "favicon.ico"
|
301 |
return FileResponse(favicon_path, media_type="image/x-icon")
|
302 |
|
303 |
@app.get("/banner.jpg")
|
304 |
+
async def banner():
|
305 |
+
banner_path = Path(__file__).parent / "banner.jpg"
|
306 |
+
return FileResponse(banner_path, media_type="image/jpeg")
|
307 |
|
308 |
@app.get("/ping")
|
309 |
async def ping():
|
|
|
315 |
if html_content is None:
|
316 |
return HTMLResponse(content="<h1>File not found</h1>", status_code=404)
|
317 |
return HTMLResponse(content=html_content)
|
318 |
+
|
319 |
@app.get("/script.js", response_class=HTMLResponse)
|
320 |
+
async def script():
|
321 |
html_content = read_html_file("script.js")
|
322 |
if html_content is None:
|
323 |
return HTMLResponse(content="<h1>File not found</h1>", status_code=404)
|
324 |
return HTMLResponse(content=html_content)
|
325 |
+
|
326 |
@app.get("/style.css", response_class=HTMLResponse)
|
327 |
+
async def style():
|
328 |
html_content = read_html_file("style.css")
|
329 |
if html_content is None:
|
330 |
return HTMLResponse(content="<h1>File not found</h1>", status_code=404)
|
331 |
return HTMLResponse(content=html_content)
|
332 |
+
|
333 |
@app.get("/dynamo", response_class=HTMLResponse)
|
334 |
async def dynamic_ai_page(request: Request):
|
335 |
user_agent = request.headers.get('user-agent', 'Unknown User')
|
336 |
client_ip = request.client.host
|
337 |
location = f"IP: {client_ip}"
|
338 |
+
|
339 |
prompt = f"""
|
340 |
+
Generate a dynamic HTML page for a user with the following details: with name "LOKI.AI"
|
341 |
- User-Agent: {user_agent}
|
342 |
- Location: {location}
|
343 |
- Style: Cyberpunk, minimalist, or retro
|
344 |
+
|
345 |
Make sure the HTML is clean and includes a heading, also have cool animations a motivational message, and a cool background.
|
346 |
Wrap the generated HTML in triple backticks (```).
|
347 |
"""
|
348 |
+
|
349 |
payload = {
|
350 |
"model": "mistral-small-latest",
|
351 |
"messages": [{"role": "user", "content": prompt}]
|
352 |
}
|
353 |
+
|
354 |
headers = {
|
355 |
"Authorization": "Bearer playground"
|
356 |
}
|
357 |
+
|
358 |
+
response = requests.post("[https://parthsadaria-lokiai.hf.space/chat/completions](https://parthsadaria-lokiai.hf.space/chat/completions)", json=payload, headers=headers)
|
359 |
data = response.json()
|
360 |
+
|
|
|
361 |
html_content = re.search(r"```(.*?)```", data['choices'][0]['message']['content'], re.DOTALL)
|
362 |
if html_content:
|
363 |
html_content = html_content.group(1).strip()
|
364 |
+
|
|
|
365 |
if html_content:
|
366 |
html_content = ' '.join(html_content.split(' ')[1:])
|
|
|
|
|
367 |
|
368 |
+
return HTMLResponse(content=html_content)
|
|
|
369 |
|
370 |
@app.get("/scraper", response_class=PlainTextResponse)
|
371 |
def scrape_site(url: str = Query(..., description="URL to scrape")):
|
372 |
try:
|
|
|
373 |
scraper = cloudscraper.create_scraper()
|
374 |
response = scraper.get(url)
|
375 |
if response.status_code == 200 and len(response.text.strip()) > 0:
|
|
|
378 |
print(f"Cloudscraper failed: {e}")
|
379 |
return "Cloudscraper failed."
|
380 |
|
|
|
|
|
|
|
381 |
@app.get("/playground", response_class=HTMLResponse)
|
382 |
async def playground():
|
383 |
html_content = read_html_file("playground.html")
|
384 |
if html_content is None:
|
385 |
return HTMLResponse(content="<h1>playground.html not found</h1>", status_code=404)
|
386 |
return HTMLResponse(content=html_content)
|
387 |
+
|
388 |
@app.get("/image-playground", response_class=HTMLResponse)
|
389 |
+
async def image_playground():
|
390 |
html_content = read_html_file("image-playground.html")
|
391 |
if html_content is None:
|
392 |
return HTMLResponse(content="<h1>image-playground.html not found</h1>", status_code=404)
|
393 |
return HTMLResponse(content=html_content)
|
394 |
|
395 |
+
GITHUB_BASE = "[https://raw.githubusercontent.com/Parthsadaria/Vetra/main](https://raw.githubusercontent.com/Parthsadaria/Vetra/main)"
|
|
|
|
|
|
|
|
|
396 |
|
397 |
FILES = {
|
398 |
"html": "index.html",
|
|
|
425 |
|
426 |
return HTMLResponse(content=final_html)
|
427 |
|
|
|
|
|
|
|
|
|
|
|
428 |
@app.get("/api/v1/models")
|
429 |
@app.get("/models")
|
430 |
async def return_models():
|
431 |
return await get_models()
|
432 |
|
|
|
433 |
@app.get("/searchgpt")
|
434 |
async def search_gpt(q: str, stream: Optional[bool] = False, systemprompt: Optional[str] = None):
|
435 |
if not q:
|
|
|
460 |
media_type="text/event-stream"
|
461 |
)
|
462 |
else:
|
|
|
463 |
collected_text = ""
|
464 |
while True:
|
465 |
item = await queue.get()
|
|
|
473 |
|
474 |
return JSONResponse(content={"response": collected_text})
|
475 |
|
|
|
|
|
|
|
476 |
header_url = os.getenv('HEADER_URL')
|
477 |
@app.post("/chat/completions")
|
478 |
@app.post("/api/v1/chat/completions")
|
479 |
async def get_completion(payload: Payload, request: Request, authenticated: bool = Depends(verify_api_key)):
|
|
|
480 |
if not server_status:
|
481 |
return JSONResponse(
|
482 |
status_code=503,
|
|
|
485 |
|
486 |
model_to_use = payload.model or "gpt-4o-mini"
|
487 |
|
|
|
488 |
if available_model_ids and model_to_use not in set(available_model_ids):
|
489 |
raise HTTPException(
|
490 |
status_code=400,
|
491 |
detail=f"Model '{model_to_use}' is not available. Check /models for the available model list."
|
492 |
)
|
493 |
|
|
|
494 |
asyncio.create_task(log_request(request, model_to_use))
|
495 |
usage_tracker.record_request(model=model_to_use, endpoint="/chat/completions")
|
496 |
|
|
|
497 |
payload_dict = payload.dict()
|
498 |
payload_dict["model"] = model_to_use
|
499 |
|
|
|
500 |
stream_enabled = payload_dict.get("stream", True)
|
501 |
|
|
|
502 |
env_vars = get_env_vars()
|
503 |
|
|
|
504 |
if model_to_use in mistral_models:
|
505 |
endpoint = env_vars['mistral_api']
|
506 |
custom_headers = {
|
|
|
512 |
elif model_to_use in alternate_models:
|
513 |
endpoint = env_vars['secret_api_endpoint_2']
|
514 |
custom_headers = {}
|
515 |
+
elif model_to_use in claude_3_models:
|
516 |
endpoint = env_vars['secret_api_endpoint_5']
|
517 |
custom_headers = {}
|
518 |
+
elif model_to_use in gemini_models: # Handle Gemini models
|
519 |
+
endpoint = env_vars['secret_api_endpoint_6']
|
520 |
+
if not endpoint:
|
521 |
+
raise HTTPException(status_code=500, detail="Gemini API endpoint not configured")
|
522 |
+
if not env_vars['gemini_key']:
|
523 |
+
raise HTTPException(status_code=500, detail="GEMINI_KEY not configured")
|
524 |
+
custom_headers = {
|
525 |
+
"Authorization": f"Bearer {env_vars['gemini_key']}"
|
526 |
+
}
|
527 |
else:
|
528 |
endpoint = env_vars['secret_api_endpoint']
|
529 |
custom_headers = {
|
|
|
534 |
|
535 |
print(f"Using endpoint: {endpoint} for model: {model_to_use}")
|
536 |
|
|
|
537 |
async def real_time_stream_generator():
|
538 |
try:
|
539 |
async with httpx.AsyncClient(timeout=60.0) as client:
|
|
|
548 |
detail = error_messages.get(response.status_code, f"Error code: {response.status_code}")
|
549 |
raise HTTPException(status_code=response.status_code, detail=detail)
|
550 |
|
|
|
551 |
async for line in response.aiter_lines():
|
552 |
if line:
|
|
|
553 |
yield line + "\n"
|
554 |
except httpx.TimeoutException:
|
555 |
raise HTTPException(status_code=504, detail="Request timed out")
|
|
|
560 |
raise e
|
561 |
raise HTTPException(status_code=500, detail=f"An error occurred: {str(e)}")
|
562 |
|
|
|
563 |
if stream_enabled:
|
564 |
return StreamingResponse(
|
565 |
real_time_stream_generator(),
|
|
|
568 |
"Content-Type": "text/event-stream",
|
569 |
"Cache-Control": "no-cache",
|
570 |
"Connection": "keep-alive",
|
571 |
+
"X-Accel-Buffering": "no"
|
572 |
}
|
573 |
)
|
574 |
else:
|
|
|
575 |
response_content = []
|
576 |
async for chunk in real_time_stream_generator():
|
577 |
response_content.append(chunk)
|
|
|
578 |
return JSONResponse(content=json.loads(''.join(response_content)))
|
579 |
|
|
|
|
|
|
|
580 |
@app.post("/images/generations")
|
581 |
async def create_image(payload: ImageGenerationPayload, authenticated: bool = Depends(verify_api_key)):
|
|
|
|
|
|
|
|
|
582 |
if not server_status:
|
583 |
return JSONResponse(
|
584 |
status_code=503,
|
585 |
content={"message": "Server is under maintenance. Please try again later."}
|
586 |
)
|
587 |
|
|
|
588 |
if payload.model not in supported_image_models:
|
589 |
raise HTTPException(
|
590 |
status_code=400,
|
591 |
+
detail=f"Model '{payload.model}' is not supported for image generation. Supported models are: {supported_image_models}"
|
592 |
)
|
593 |
|
|
|
594 |
usage_tracker.record_request(model=payload.model, endpoint="/images/generations")
|
595 |
|
|
|
596 |
api_payload = {
|
597 |
"model": payload.model,
|
598 |
"prompt": payload.prompt,
|
|
|
600 |
"number": payload.number
|
601 |
}
|
602 |
|
|
|
603 |
target_api_url = os.getenv('NEW_IMG')
|
604 |
|
605 |
try:
|
|
|
606 |
async with httpx.AsyncClient(timeout=60.0) as client:
|
607 |
response = await client.post(target_api_url, json=api_payload)
|
608 |
|
|
|
610 |
error_detail = response.json().get("detail", f"Image generation failed with status code: {response.status_code}")
|
611 |
raise HTTPException(status_code=response.status_code, detail=error_detail)
|
612 |
|
|
|
613 |
return JSONResponse(content=response.json())
|
614 |
|
615 |
except httpx.TimeoutException:
|
|
|
619 |
except Exception as e:
|
620 |
raise HTTPException(status_code=500, detail=f"An unexpected error occurred during image generation: {e}")
|
621 |
|
|
|
|
|
|
|
622 |
async def log_request(request, model):
|
|
|
623 |
current_time = (datetime.datetime.utcnow() + datetime.timedelta(hours=5, minutes=30)).strftime("%Y-%m-%d %I:%M:%S %p")
|
624 |
+
ip_hash = hash(request.client.host) % 10000
|
625 |
print(f"Time: {current_time}, IP Hash: {ip_hash}, Model: {model}")
|
626 |
|
|
|
627 |
@lru_cache(maxsize=10)
|
628 |
def get_usage_summary(days=7):
|
629 |
return usage_tracker.get_usage_summary(days)
|
630 |
|
631 |
@app.get("/usage")
|
632 |
async def get_usage(days: int = 7):
|
|
|
633 |
return get_usage_summary(days)
|
634 |
|
|
|
635 |
def generate_usage_html(usage_data):
|
|
|
636 |
model_usage_rows = "\n".join([
|
637 |
f"""
|
638 |
<tr>
|
|
|
644 |
""" for model, model_data in usage_data['models'].items()
|
645 |
])
|
646 |
|
|
|
647 |
api_usage_rows = "\n".join([
|
648 |
f"""
|
649 |
<tr>
|
|
|
655 |
""" for endpoint, endpoint_data in usage_data['api_endpoints'].items()
|
656 |
])
|
657 |
|
|
|
658 |
daily_usage_rows = "\n".join([
|
659 |
"\n".join([
|
660 |
f"""
|
|
|
673 |
<head>
|
674 |
<meta charset="UTF-8">
|
675 |
<title>Lokiai AI - Usage Statistics</title>
|
676 |
+
<link href="[https://fonts.googleapis.com/css2?family=Inter:wght@300;400;600&display=swap](https://fonts.googleapis.com/css2?family=Inter:wght@300;400;600&display=swap)" rel="stylesheet">
|
677 |
<style>
|
678 |
:root {{
|
679 |
--bg-dark: #0f1011;
|
|
|
819 |
"""
|
820 |
return html_content
|
821 |
|
|
|
822 |
@lru_cache(maxsize=1)
|
823 |
def get_usage_page_html():
|
824 |
usage_data = get_usage_summary()
|
|
|
826 |
|
827 |
@app.get("/usage/page", response_class=HTMLResponse)
|
828 |
async def usage_page():
|
|
|
|
|
829 |
html_content = get_usage_page_html()
|
830 |
return HTMLResponse(content=html_content)
|
831 |
|
|
|
832 |
@app.get("/meme")
|
833 |
async def get_meme():
|
834 |
try:
|
|
|
835 |
client = get_async_client()
|
836 |
+
response = await client.get("[https://meme-api.com/gimme](https://meme-api.com/gimme)")
|
837 |
response_data = response.json()
|
838 |
|
839 |
meme_url = response_data.get("url")
|
|
|
842 |
|
843 |
image_response = await client.get(meme_url, follow_redirects=True)
|
844 |
|
|
|
845 |
async def stream_with_larger_chunks():
|
846 |
chunks = []
|
847 |
size = 0
|
848 |
async for chunk in image_response.aiter_bytes(chunk_size=16384):
|
849 |
chunks.append(chunk)
|
850 |
size += len(chunk)
|
|
|
851 |
if size >= 65536:
|
852 |
yield b''.join(chunks)
|
853 |
chunks = []
|
854 |
size = 0
|
|
|
855 |
if chunks:
|
856 |
yield b''.join(chunks)
|
857 |
|
858 |
return StreamingResponse(
|
859 |
stream_with_larger_chunks(),
|
860 |
media_type=image_response.headers.get("content-type", "image/png"),
|
861 |
+
headers={'Cache-Control': 'max-age=3600'}
|
862 |
)
|
863 |
except Exception:
|
864 |
raise HTTPException(status_code=500, detail="Failed to retrieve meme")
|
865 |
|
|
|
866 |
def load_model_ids(json_file_path):
|
867 |
try:
|
868 |
with open(json_file_path, 'r') as f:
|
869 |
models_data = json.load(f)
|
|
|
870 |
return [model['id'] for model in models_data if 'id' in model]
|
871 |
except Exception as e:
|
872 |
print(f"Error loading model IDs: {str(e)}")
|
|
|
878 |
available_model_ids = load_model_ids("models.json")
|
879 |
print(f"Loaded {len(available_model_ids)} model IDs")
|
880 |
|
|
|
881 |
available_model_ids.extend(list(pollinations_models))
|
|
|
882 |
available_model_ids.extend(list(alternate_models))
|
|
|
883 |
available_model_ids.extend(list(mistral_models))
|
|
|
884 |
available_model_ids.extend(list(claude_3_models))
|
885 |
+
available_model_ids.extend(list(gemini_models)) # Add Gemini models
|
886 |
|
887 |
+
available_model_ids = list(set(available_model_ids))
|
888 |
print(f"Total available models: {len(available_model_ids)}")
|
889 |
|
|
|
890 |
for _ in range(MAX_SCRAPERS):
|
891 |
scraper_pool.append(cloudscraper.create_scraper())
|
892 |
|
|
|
893 |
env_vars = get_env_vars()
|
894 |
missing_vars = []
|
895 |
|
|
|
903 |
missing_vars.append('SECRET_API_ENDPOINT_3')
|
904 |
if not env_vars['secret_api_endpoint_4']:
|
905 |
missing_vars.append('SECRET_API_ENDPOINT_4')
|
906 |
+
if not env_vars['secret_api_endpoint_5']:
|
907 |
missing_vars.append('SECRET_API_ENDPOINT_5')
|
908 |
+
if not env_vars['secret_api_endpoint_6']: # Check the new endpoint
|
909 |
+
missing_vars.append('SECRET_API_ENDPOINT_6')
|
910 |
if not env_vars['mistral_api'] and any(model in mistral_models for model in available_model_ids):
|
911 |
missing_vars.append('MISTRAL_API')
|
912 |
if not env_vars['mistral_key'] and any(model in mistral_models for model in available_model_ids):
|
913 |
missing_vars.append('MISTRAL_KEY')
|
914 |
+
if not env_vars['gemini_key'] and any(model in gemini_models for model in available_model_ids): # Check Gemini key
|
915 |
+
missing_vars.append('GEMINI_KEY')
|
916 |
|
917 |
if missing_vars:
|
918 |
print(f"WARNING: The following environment variables are missing: {', '.join(missing_vars)}")
|
|
|
922 |
|
923 |
@app.on_event("shutdown")
|
924 |
async def shutdown_event():
|
|
|
925 |
client = get_async_client()
|
926 |
await client.aclose()
|
|
|
|
|
927 |
scraper_pool.clear()
|
|
|
|
|
928 |
usage_tracker.save_data()
|
|
|
929 |
print("Server shutdown complete!")
|
930 |
|
|
|
|
|
931 |
@app.get("/health")
|
932 |
async def health_check():
|
|
|
933 |
env_vars = get_env_vars()
|
934 |
missing_critical_vars = []
|
935 |
|
|
|
936 |
if not env_vars['api_keys'] or env_vars['api_keys'] == ['']:
|
937 |
missing_critical_vars.append('API_KEYS')
|
938 |
if not env_vars['secret_api_endpoint']:
|
|
|
943 |
missing_critical_vars.append('SECRET_API_ENDPOINT_3')
|
944 |
if not env_vars['secret_api_endpoint_4']:
|
945 |
missing_critical_vars.append('SECRET_API_ENDPOINT_4')
|
946 |
+
if not env_vars['secret_api_endpoint_5']:
|
947 |
missing_critical_vars.append('SECRET_API_ENDPOINT_5')
|
948 |
+
if not env_vars['secret_api_endpoint_6']: # Check the new endpoint
|
949 |
+
missing_critical_vars.append('SECRET_API_ENDPOINT_6')
|
950 |
if not env_vars['mistral_api']:
|
951 |
missing_critical_vars.append('MISTRAL_API')
|
952 |
if not env_vars['mistral_key']:
|
953 |
missing_critical_vars.append('MISTRAL_KEY')
|
954 |
+
if not env_vars['gemini_key']: # Check Gemini key
|
955 |
+
missing_critical_vars.append('GEMINI_KEY')
|
956 |
|
957 |
health_status = {
|
958 |
"status": "healthy" if not missing_critical_vars else "unhealthy",
|
|
|
964 |
|
965 |
if __name__ == "__main__":
|
966 |
import uvicorn
|
967 |
+
uvicorn.run(app, host="0.0.0.0", port=7860)
|