Velocity-Adaptive-Guidance-Scale/VAGS_Generation/evaluation_sd35_coco17.sh at main · Harvard-AI-and-Robotics-Lab/Velocity-Adaptive-Guidance-Scale · 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
#!/bin/bash

source activate base
conda activate vags_gen

HUGGINGFACE_TOKEN=''
BATCH_SIZE=2
NUM_PROMPTS_PER_RUN=4
NUM_PROMPTS_PER_RUN=1
IMAGE_WIDTH=512
IMAGE_HEIGHT=512

SCHEDULER="flow_euler" #"flowmatch_euler"
GUIDANCE=( 3.5 )
INFERENCE_STEPS=28

ORG_IMAGE_DIR="/path/to/datasets/coco17/validation_bestcaption"
PROMPT_DIR="/path/to/datasets/coco17/validation_bestcaption"
# BASE_OUTPUT_DIR="generated_images_coco17_bestcaption"

OUTPUT_DIRS=(
            # 'outputs/coco17_sd35_g3p5_step28_seed42/vagsgen_k0'
            # 'outputs/coco17_sd35_g3p5_step28_seed42/vagsgen_k0p1'
            # 'outputs/coco17_sd35_g3p5_step28_seed42/vagsgen_k0p3'
            # 'outputs/coco17_sd35_g3p5_step28_seed42/vagsgen_k0p5'
            # 'outputs/coco17_sd35_g3p5_step28_seed42/vagsgen_k0p7'
            # 'outputs/coco17_sd35_g3p5_step28_seed42/vagsgen_k0p9'
            # 'outputs/coco17_sd35/vagsgen_k0p5'
            # 'outputs/coco17_sd35/vagsgen_k0p7'
            # 'outputs/coco17_sd35/vagsgen_k0p9'
            # 'outputs/coco17_sd35/vagsgen_k1'
            outputs/coco17_sd35/vagsgen_k0
            outputs/coco17_sd35/vagsgen_k0p1
            outputs/coco17_sd35/vagsgen_k0p3
            outputs/coco17_sd35/vagsgen_k1p5
            outputs/coco17_sd35/vagsgen_k2
            outputs/coco17_sd35/vagsgen_k3
            outputs/coco17_sd35/vagsgen_k4
            outputs/coco17_sd35/vagsgen_k5
            )


for (( i=0; i<${#OUTPUT_DIRS[@]}; i++ ));
do

# OUTPUT_DIR="${BASE_OUTPUT_DIR}_${SCHEDULER}_step${INFERENCE_STEPS}_g${GUIDANCE[$i]}"
# GEN_IMG_DIR="${MODEL_PATH}/${OUTPUT_DIR}"

# python inference_sd35.py \
#     --prompt_dir=${PROMPT_DIR} \
#     --output_dir=${OUTPUT_DIRS[$i]} \
#     --batch_size=${BATCH_SIZE} \
#     --num_prompts_per_run=${NUM_PROMPTS_PER_RUN} \
#     --image_width=${IMAGE_WIDTH} \
#     --image_height=${IMAGE_HEIGHT} \
#     --num_inference_steps=${INFERENCE_STEPS} \
#     --guidance_scale=${GUIDANCE} \
#     --scheduler=${SCHEDULER} \
#     --seed 42
# # --model_path=${MODEL_PATH} \

python evaluate_on_image_generation.py \
    --prompt_dir=${PROMPT_DIR} \
    --gt_img_dir=${ORG_IMAGE_DIR} \
    --gen_img_dir=${OUTPUT_DIRS[$i]}

python evaluate_with_blip_caption.py \
    --caption_folder=${PROMPT_DIR} \
    --image_folder=${OUTPUT_DIRS[$i]}

parent=$(dirname "${OUTPUT_DIRS[$i]}")
current_dir=$(basename "${OUTPUT_DIRS[$i]}")
python assemble_performance_scores.py \
    --csv_file="${parent}/${current_dir}_img_quality.txt" \
    --json_file="${parent}/${current_dir}_caption_results.json" \
    --output="${parent}/${current_dir}_perf.txt"
done