project_setter/anima_module.yaml at main · RobotFlow-Labs/project_setter · GitHub

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
schema: anima.module/v1

module:
  name: project_setter
  codename: SETTER
  wave: 10
  theme: WARDOG
  tier: T2
  score: 37

paper:
  title: "SFFNet: Synergistic Feature Fusion for UAV Image Detection"
  arxiv_id: "2604.03176"
  arxiv_url: https://arxiv.org/abs/2604.03176
  github: https://github.com/CQNU-ZhangLab/SFFNet
  status: partial  # upstream code is incomplete; we reimplement from paper

backbone:
  framework: ultralytics
  model: yolo26m
  custom_yaml: configs/models/yolo26m-sff.yaml

compute:
  backends: [cuda, mlx, cpu]
  default: cuda
  min_vram_gb: 14
  target_vram_gb: 18

datasets:
  provenance: configs/datasets/provenance.json
  required:
    - VisDrone
  optional:
    - UAVDT
    - SeaDronesSee
    - UAVid
    - MegaUAV-1.8M
    - NighthawkMegaHiRes

artifacts:
  project_dir: /mnt/artifacts-datai/project_setter
  checkpoints: /mnt/artifacts-datai/checkpoints/project_setter
  logs: /mnt/artifacts-datai/logs/project_setter
  tensorboard: /mnt/artifacts-datai/tensorboard/project_setter
  exports: /mnt/artifacts-datai/exports/project_setter

exports:
  formats: [pth, safetensors, onnx, trt_fp16, trt_fp32]
  hf_repo: ilessio-aiflowlab/project_setter

docker:
  cuda:
    dockerfile: docker/Dockerfile.cuda
    image: robotflowlabs/anima-setter:cuda
  mlx:
    dockerfile: docker/Dockerfile.mlx
    image: robotflowlabs/anima-setter:mlx

serve:
  entrypoint: anima_setter.serve:build_app
  port: 8080
  env:
    ANIMA_BACKEND: cuda
    ANIMA_MODEL_PATH: /models/setter.pt
  healthcheck: GET /health

ros2:
  contract: src/anima_setter/ros2_contract.py
  topics:
    input:
      - name: /setter/image_in
        type: sensor_msgs/Image
    output:
      - name: /setter/detections
        type: setter_msgs/DetectionArray
      - name: /setter/status
        type: setter_msgs/Status

entrypoints:
  cli: anima_setter.cli:main
  train: scripts/train_sff.py
  preflight: scripts/preflight_datasets.py

siblings:
  count: 33
  wave_theme: "War Dog Breeds — UAV/Drone defense"