File size: 2,105 Bytes
4336727
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
# Low Light Rainy Code Export

This folder is a portable code-only export of `Low_light_rainy_new`.

## Excluded

The following large/runtime folders were intentionally excluded:

- `dataset/`
- `checkpoint*/`
- `results/`
- Python caches
- `.pth`, `.zip`, `.log` files

## Included Core Code

Key files for the current restoration experiments:

- `train_restormer.py` - main trainer for baseline, v1, and v2.
- `restormer.py` - original Restormer backbone and existing variants.
- `net/restormer_lowlight_rain.py` - FGP-Restormer v1.
- `net/restormer_lowlight_rain_v2.py` - FGP-Restormer v2.
- `utils/inference_utils.py` - validation padding, crop, RGB PSNR, optional TTA.
- `run_restormer_fgp.sh` - v1 launch script.
- `run_restormer_fgp_v2.sh` - v2 launch script.
- `dataset_RGB_ori.py`, `data_RGB.py` - legacy data loaders.
- `dataset_rwjr10k.py`, `train_restormer_rwjr10k.py`, `train_restormer_rwjr10k_from_original.py` - RWJR-10K related code, if needed.

## Environment

The current `new` conda environment was exported to:

- `requirements_new.txt`
- `environment_new.yml`

Recommended setup on a new server:

```bash
conda env create -f environment_new.yml
conda activate new
```

If the environment name conflicts, edit the first line of `environment_new.yml`, or use pip inside your own environment:

```bash
python -m pip install -r requirements_new.txt
```

## Data Layout Expected by Current v1/v2 Scripts

Default low-light-rain training expects:

```text
dataset/train/syn+real/input
dataset/train/syn+real/target
dataset/test/input
dataset/test/target
```

The v1/v2 scripts do not include dataset files in this export. Copy datasets separately to the same relative locations or pass explicit paths:

```bash
python train_restormer.py \
  --train_inp /path/to/input \
  --train_tar /path/to/target \
  --test_inp /path/to/test/input \
  --test_tar /path/to/test/target
```

## Run v1

```bash
bash run_restormer_fgp.sh
```

## Run v2

```bash
bash run_restormer_fgp_v2.sh
```

Before running on a new server, update any `--resume` paths in the run scripts if checkpoints are not present.