[Python] How do I deploy local model weights (best.pt) to CVAT for auto-annotation

I'm trying to use CVAT locally with Docker for image bounding box labeling. To save time, I'd like to use auto-annotation with a partially trained model. However, when I try to deploy the model from WSL, I keep getting an error saying requirements.txt is missing. I'm not sure how to proceed. Any help would be greatly appreciated.

@DESKTOP-HS1UL5Q:~/.nuclio/best-model$ nuctl deploy --project-name cvat --path . --platform local


enter image description here

I'm trying to deploy the model by running the code above, but it can't find requirements.txt. The image below shows the file structure.

enter image description here

Here are the contents of each file.

function.yaml

metadata:
name: best-model
namespace: cvat
spec:
runtime: python:3.9
handler: handler:handler
description: YOLOv8 object detector for CVAT
env: []
build:
commands:
- pip install -r requirements.txt
artifacts:
- target: best.pt
- target: handler.py
- target: requirements.txt
triggers:
http:
kind: http
maxWorkers: 1
workerAvailabilityTimeoutMilliseconds: 10000
attributes:
maxRequestBodySize: 33554432


handler.py

import io
import json
import numpy as np
import cv2
from PIL import Image
import base64
from ultralytics import YOLO

model = YOLO("best.pt")

def handler(context, event):
body = event.body
img_bytes = np.asarray(bytearray(body), dtype=np.uint8)
img = cv2.imdecode(img_bytes, cv2.IMREAD_COLOR)

results = model(img)[0]
predictions = []

for box in results.boxes:
cls = int(box.cls[0].item())
conf = float(box.conf[0].item())
x1, y1, x2, y2 = map(float, box.xyxy[0].tolist())

predictions.append({
"confidence": conf,
"label": str(cls),
"points": [[x1, y1], [x2, y2]],
"type": "rectangle"
})

return context.Response(body=json.dumps(predictions), headers={},
content_type='application/json', status_code=200)


requirements.txt

ultralytics==8.1.24
opencv-python
numpy

Continue reading...