forked from roboflow/inference
-
Notifications
You must be signed in to change notification settings - Fork 0
Expand file tree
/
Copy pathdebugrun.py
More file actions
52 lines (42 loc) · 1.62 KB
/
debugrun.py
File metadata and controls
52 lines (42 loc) · 1.62 KB
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
import sys
import uvicorn
from inference.core.cache import cache
from inference.core.env import MAX_ACTIVE_MODELS, ENABLE_STREAM_API
from inference.models.utils import ROBOFLOW_MODEL_TYPES
from inference.core.interfaces.http.http_api import HttpInterface
from inference.core.managers.active_learning import BackgroundTaskActiveLearningManager
from inference.core.managers.decorators.fixed_size_cache import WithFixedSizeCache
from inference.core.registries.roboflow import RoboflowModelRegistry
from inference.core.interfaces.stream_manager.manager_app.app import start
"""
convenient script to run server in debug
(i.e. runs uvicorn directly)
It's a simplified version of docker/config/cpu_http.py
see https://www.loom.com/share/48f71894427a473cac39eca25f6ac759
- uv venv
- source .venv/bin/activate
- uv pip install -e .
- # start debugrun.py in debug mode
"""
model_registry = RoboflowModelRegistry(ROBOFLOW_MODEL_TYPES)
model_manager = BackgroundTaskActiveLearningManager(
model_registry=model_registry, cache=cache
)
model_manager = WithFixedSizeCache(model_manager, max_size=MAX_ACTIVE_MODELS)
model_manager.init_pingback()
interface = HttpInterface(model_manager)
app = interface.app
from functools import partial
from multiprocessing import Process
if __name__ == "__main__":
print("Starting Stream Manager...")
stream_manager_process = Process(
target=partial
36A7
(start, expected_warmed_up_pipelines=0),
)
stream_manager_process.start()
print("Starting server...")
try:
uvicorn.run(app, host="0.0.0.0", port=9001)
except Exception as e:
print("Error starting server:", e)
sys.exit(1)