File size: 1,772 Bytes
5b4786e
 
 
8b2bb9b
54947ac
 
 
 
 
5b4786e
 
a5cd004
1f30fb0
 
 
 
5b4786e
 
 
 
8b2bb9b
 
 
e4249e7
8b2bb9b
 
 
fca2c62
54947ac
fca2c62
a5cd004
 
fca2c62
 
54947ac
5b4786e
 
 
a5cd004
5b4786e
fca2c62
 
54947ac
fca2c62
5b4786e
fca2c62
 
54947ac
5b4786e
 
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
import numpy as np
import torch
from transformers import TimesformerForVideoClassification
from preprocessing import read_video
import logging

# ๋กœ๊น… ์„ค์ •
logging.basicConfig(level=logging.INFO)
logger = logging.getLogger(__name__)

class EndpointHandler:
    def __init__(self, model_dir):
        self.model = TimesformerForVideoClassification.from_pretrained(
            'donghuna/timesformer-base-finetuned-k400-diving48',
            ignore_mismatched_sizes=True
        )
        self.model.classifier = torch.nn.Linear(self.model.classifier.in_features, 48)  # 48 output classes
        self.model.eval()
        
    def __call__(self, data):
        video_path = data["inputs"]["video_path"]
        ftp_password = data["inputs"].get("ftp_password")

        processed_frames = read_video(video_path, ftp_password)


        
        # ๋””๋ฒ„๊น…: ์ž…๋ ฅ ๋ฐ์ดํ„ฐ ํ™•์ธ
        logger.info(f"Received data: {data}")

        frames = np.array(data['frames'])
        frames = torch.tensor(frames).float()  # Ensure the data is in the correct format

        # ๋””๋ฒ„๊น…: ํ”„๋ ˆ์ž„ ๋ฐ์ดํ„ฐ ํ™•์ธ
        logger.info(f"Frames shape: {frames.shape}")
        
        # Perform inference
        with torch.no_grad():
            outputs = self.model(frames.unsqueeze(0))  # Add batch dimension
            predictions = torch.softmax(outputs.logits, dim=-1)

            # ๋””๋ฒ„๊น…: ์˜ˆ์ธก ๊ฒฐ๊ณผ ํ™•์ธ
            logger.info(f"Predictions: {predictions}")
            
            predicted_class = torch.argmax(predictions, dim=-1).item()

            # ๋””๋ฒ„๊น…: ์˜ˆ์ธก ํด๋ž˜์Šค ํ™•์ธ
            logger.info(f"Predicted class: {predicted_class}")
        
        return {"predicted_class": predicted_class, "predictions": predictions.tolist()}