This repository was archived by the owner on May 13, 2025. It is now read-only.
-
Notifications
You must be signed in to change notification settings - Fork 0
Expand file tree
/
Copy pathtest_v2.py
More file actions
130 lines (105 loc) · 8.8 KB
/
Copy pathtest_v2.py
File metadata and controls
130 lines (105 loc) · 8.8 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
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
"""
Temp functional test code
"""
import os
import gc
import sys
import multiprocessing
from typing import Tuple, Mapping, Any, List, Dict
sys.path.append(os.path.join(os.path.dirname(os.getcwd()), "../../"))
import torch
from tqdm import tqdm
from torch.nn import Module
from AI.src.utils import Logger
from AI.src.data.dataloader import DefaultDataLoader
from AI.src.utils.inference_ops import dispatch_infer
from AI.src.modeling.architectures import build_model
from AI.src.data.dataset import VADFrameLevelTestDataset
from AI.src.utils import DotDict, load_config, load_weights, VideoCache
global starter
def main() -> None:
multiprocessing.set_start_method("spawn")
device = "cuda"
T_max: int = 30
overlap_ratio: float = 0.5
batch_thres = [1000, 2000, 3000, 4500, 6000, 10000, 15000, 20000, torch.inf]
batch_worker = [16] * 4 + [14] * 2 + [12] * 3
video_cache = VideoCache(batch_thres, batch_worker)
for pred_result, config, weight in zip(
[
"/home/trong/Downloads/Local/Source/Python/semester_9/AIP391/Video_anomaly_detection/AI/results/final_train_result/teacher/input_2d/v1/log/pred_result.txt",
# "/home/trong/Downloads/Local/Source/Python/semester_9/AIP391/Video_anomaly_detection/AI/results/final_train_result/teacher/input_2d/v2/log/pred_result.txt",
# "/home/trong/Downloads/Local/Source/Python/semester_9/AIP391/Video_anomaly_detection/AI/results/final_train_result/teacher/input_2d/v3/log/pred_result.txt",
# "/home/trong/Downloads/Local/Source/Python/semester_9/AIP391/Video_anomaly_detection/AI/results/final_train_result/teacher/input_3d/v1/log/pred_result.txt",
"/home/trong/Downloads/Local/Source/Python/semester_9/AIP391/Video_anomaly_detection/AI/results/final_train_result/teacher/input_3d/v2/log/pred_result.txt",
# "/home/trong/Downloads/Local/Source/Python/semester_9/AIP391/Video_anomaly_detection/AI/results/final_train_result/teacher/input_3d/v3/log/pred_result.txt",
# "/home/trong/Downloads/Local/Source/Python/semester_9/AIP391/Video_anomaly_detection/AI/results/final_train_result/teacher/input_mixed/v1/log/pred_result.txt",
# "/home/trong/Downloads/Local/Source/Python/semester_9/AIP391/Video_anomaly_detection/AI/results/final_train_result/teacher/input_mixed/v2/log/pred_result.txt",
"/home/trong/Downloads/Local/Source/Python/semester_9/AIP391/Video_anomaly_detection/AI/results/final_train_result/teacher/input_mixed/v3/log/pred_result.txt"
],
[
"/home/trong/Downloads/Local/Source/Python/semester_9/AIP391/Video_anomaly_detection/AI/results/final_train_result/teacher/input_2d/v1/Mlflow/995263845449942640/d4e6cc59499a4abc90cf6410eb9aef25/artifacts/config.json",
# "/home/trong/Downloads/Local/Source/Python/semester_9/AIP391/Video_anomaly_detection/AI/results/final_train_result/teacher/input_2d/v2/Mlflow/371379892464714510/ca8300cd05d4495db1070b9d9ea987d8/artifacts/config.json",
# "/home/trong/Downloads/Local/Source/Python/semester_9/AIP391/Video_anomaly_detection/AI/results/final_train_result/teacher/input_2d/v3/Mlflow/118746316556651065/dd5c8e46f4274cd481ea91465b591ea9/artifacts/config.json",
# "/home/trong/Downloads/Local/Source/Python/semester_9/AIP391/Video_anomaly_detection/AI/results/final_train_result/teacher/input_3d/v1/Mlflow/890005033899140195/6653ca3fc5da40cbaf0101124bac5d70/artifacts/config.json",
"/home/trong/Downloads/Local/Source/Python/semester_9/AIP391/Video_anomaly_detection/AI/results/final_train_result/teacher/input_3d/v2/Mlflow/341932603411297071/49d947ac8c4c43758a91dbbfb4b1505c/artifacts/config.json",
# "/home/trong/Downloads/Local/Source/Python/semester_9/AIP391/Video_anomaly_detection/AI/results/final_train_result/teacher/input_3d/v3/Mlflow/510859847621159024/b814d78b360d4a09a29847d7f65d44e6/artifacts/config.json",
# "/home/trong/Downloads/Local/Source/Python/semester_9/AIP391/Video_anomaly_detection/AI/results/final_train_result/teacher/input_mixed/v1/Mlflow/401041024666025539/619caa95c3a4476b89d5d91b0fc59cc1/artifacts/config.json",
# "/home/trong/Downloads/Local/Source/Python/semester_9/AIP391/Video_anomaly_detection/AI/results/final_train_result/teacher/input_mixed/v2/Mlflow/825420379933310267/7b53e9cc5e3c4835b086a18e76d0063d/artifacts/config.json",
"/home/trong/Downloads/Local/Source/Python/semester_9/AIP391/Video_anomaly_detection/AI/results/final_train_result/teacher/input_mixed/v3/Mlflow/991378426196781897/aaff4465b324425b91421ee50a683603/artifacts/config.json"
],
[
"/home/trong/Downloads/Local/Source/Python/semester_9/AIP391/Video_anomaly_detection/AI/results/final_train_result/teacher/input_2d/v1/Mlflow/995263845449942640/d4e6cc59499a4abc90cf6410eb9aef25/artifacts/ckpt/best_epoch18_step4067.pt",
# "/home/trong/Downloads/Local/Source/Python/semester_9/AIP391/Video_anomaly_detection/AI/results/final_train_result/teacher/input_2d/v2/Mlflow/371379892464714510/ca8300cd05d4495db1070b9d9ea987d8/artifacts/ckpt/best_epoch13_step2937.pt",
# "/home/trong/Downloads/Local/Source/Python/semester_9/AIP391/Video_anomaly_detection/AI/results/final_train_result/teacher/input_2d/v3/Mlflow/118746316556651065/dd5c8e46f4274cd481ea91465b591ea9/artifacts/ckpt/best_epoch13_step2937.pt",
# "/home/trong/Downloads/Local/Source/Python/semester_9/AIP391/Video_anomaly_detection/AI/results/final_train_result/teacher/input_3d/v1/Mlflow/890005033899140195/6653ca3fc5da40cbaf0101124bac5d70/artifacts/ckpt/best_epoch14_step3163.pt",
"/home/trong/Downloads/Local/Source/Python/semester_9/AIP391/Video_anomaly_detection/AI/results/final_train_result/teacher/input_3d/v2/Mlflow/341932603411297071/49d947ac8c4c43758a91dbbfb4b1505c/artifacts/ckpt/best_epoch20_step4519.pt",
# "/home/trong/Downloads/Local/Source/Python/semester_9/AIP391/Video_anomaly_detection/AI/results/final_train_result/teacher/input_3d/v3/Mlflow/510859847621159024/b814d78b360d4a09a29847d7f65d44e6/artifacts/ckpt/best_epoch20_step4519.pt",
# "/home/trong/Downloads/Local/Source/Python/semester_9/AIP391/Video_anomaly_detection/AI/results/final_train_result/teacher/input_mixed/v1/Mlflow/401041024666025539/619caa95c3a4476b89d5d91b0fc59cc1/artifacts/ckpt/best_epoch13_step2937.pt",
# "/home/trong/Downloads/Local/Source/Python/semester_9/AIP391/Video_anomaly_detection/AI/results/final_train_result/teacher/input_mixed/v2/Mlflow/825420379933310267/7b53e9cc5e3c4835b086a18e76d0063d/artifacts/ckpt/best_epoch9_step2033.pt",
"/home/trong/Downloads/Local/Source/Python/semester_9/AIP391/Video_anomaly_detection/AI/results/final_train_result/teacher/input_mixed/v3/Mlflow/991378426196781897/aaff4465b324425b91421ee50a683603/artifacts/ckpt/best_epoch12_step2711.pt",
]
):
config: str
weight: str
logger: Logger = Logger("test")
config: DotDict = DotDict(load_config(config))
weight: Mapping[str, Any] = load_weights(weight, weights_only=False)
model: Module = build_model(config)
model.load_state_dict(weight["model"]) if isinstance(weight["model"], dict) else model.load_state_dict(
weight["model"].state_dict())
model = model.to(device)
dl = DefaultDataLoader(
VADFrameLevelTestDataset(
"/home/trong/Downloads/Dataset/VAD/final/test",
"label.csv",
), num_workers=4, shuffle=False, multiprocessing_context="fork"
)
for idx, inp, label in tqdm(dl, total=len(dl)):
idx: torch.Tensor
inp: Tuple[str]
idx: int = idx.item()
inp: str = inp[0]
video_cache.cache(label.squeeze(0).shape[0], ["inp", "label", "idx"], [inp, label, idx])
cache: Dict[str, Any] | None = video_cache.get_cache()
if cache is not None:
print("Batch:", cache["batch_worker"])
result: Tuple[List[float], List[int]] = dispatch_infer(cache, model, device, T_max, overlap_ratio)
for i in range(len(result)):
log_info = f"{result[i][0]}; {result[i][1]}; {cache['idx'][i]}\n"
logger.write(pred_result, log_info, "a")
gc.collect()
torch.cuda.empty_cache()
# Remaining
print("Run leftovers")
for cache in video_cache.get_remains():
print("Batch:", cache["batch_worker"])
result: Tuple[List[float], List[int]] = dispatch_infer(cache, model, device, T_max, overlap_ratio)
for i in range(len(result)):
log_info = f"{result[i][0]}; {result[i][1]}; {cache['idx'][i]}\n"
logger.write(pred_result, log_info, "a")
gc.collect()
torch.cuda.empty_cache()
return None
if __name__ == '__main__':
main()