features.sh

#!/bin/bash

TOTAL=1
declare -a ENCODERS=(
                   "RN50"
                #    "ViT-B/16"
                )
TOTAL=$(( TOTAL * ${#ENCODERS[@]} ))

declare -a IMAGE_LAYER_IDX=(
                   "0"
                   "1"
                )
TOTAL=$(( TOTAL * ${#IMAGE_LAYER_IDX[@]} ))

declare -a TEXT_LAYER_IDX=(
                   "0"
                #    "1"
                )
TOTAL=$(( TOTAL * ${#TEXT_LAYER_IDX[@]} ))

declare -a TEXT_AUGS=(
                    #   "classname"
                      "hand_crafted" 
                    #   "vanilla"
                    #   "template_mining"
                     )
TOTAL=$(( TOTAL * ${#TEXT_AUGS[@]} ))

declare -a IMAGE_AUGS=(
                      "none"
                      "flip" 
                    #   "randomcrop"
                     )
TOTAL=$(( TOTAL * ${#IMAGE_AUGS[@]} ))

declare -a IMAGE_VIEWS=(
                      "1"
                    #   "10"
                     )
TOTAL=$(( TOTAL * ${#IMAGE_VIEWS[@]} ))


declare -a DATASETS=(
                     "imagenet"
                     "caltech101"
                     "dtd"
                     "eurosat"
                     "fgvc_aircraft"
                     "food101"
                     "oxford_flowers"
                     "oxford_pets"
                     "stanford_cars"
                     "sun397"
                     "ucf101"
                     )
TOTAL=$(( TOTAL * ${#DATASETS[@]} ))

declare -a ALL_SHOTS=(
    "1"
    "2"
    "4"
    "8"
    "16"
)
TOTAL=$(( TOTAL * ${#ALL_SHOTS[@]} ))

declare -a ALL_SEEDS=(
    "1"
    "2"
    "3"
    # "4"
    # "5"
)
TOTAL=$(( TOTAL * ${#ALL_SEEDS[@]} ))

echo "ENCODERS: ${ENCODERS[@]}"
echo "IMAGE_LAYER_IDX: ${IMAGE_LAYER_IDX[@]}"
echo "TEXT_LAYER_IDX: ${TEXT_LAYER_IDX[@]}"
echo "TEXT_AUGS: ${TEXT_AUGS[@]}"
echo "IMAGE_AUGS: ${IMAGE_AUGS[@]}"
echo "IMAGE_VIEWS: ${IMAGE_VIEWS[@]}"
echo "DATASETS: ${DATASETS[@]}"
echo "ALL_SHOTS: ${ALL_SHOTS[@]}"
echo "ALL_SEEDS: ${ALL_SEEDS[@]}"
echo "TOTAL: $TOTAL"

COUNTER=1
echo " "
for DATASET in "${DATASETS[@]}"
do  
    echo "DATASET: $DATASET"
    for IMAGE_LAYER in "${IMAGE_LAYER_IDX[@]}"
    do 
        echo "IMAGE_LAYER: $IMAGE_LAYER"
        for ENCODER in "${ENCODERS[@]}"
        do
            echo "ENCODER: $ENCODER"
            for TEXT_LAYER in "${TEXT_LAYER_IDX[@]}"
            do
                echo "TEXT_LAYER: $TEXT_LAYER"
                for TEXT_AUG in "${TEXT_AUGS[@]}"
                do
                    echo "TEXT_AUG: $TEXT_AUG"
                    for IMAGE_AUG in "${IMAGE_AUGS[@]}"
                    do
                        echo "IMAGE_AUG: $IMAGE_AUG"
                        for IMAGE_VIEW in "${IMAGE_VIEWS[@]}"
                        do
                            echo "IMAGE_VIEW: $IMAGE_VIEW"
                            for SHOTS in "${ALL_SHOTS[@]}"
                            do 
                                echo "SHOTS: $SHOTS"
                                for SEED in "${ALL_SEEDS[@]}"
                                do
                                    echo "SEED: $SEED"
                                    echo "COUNTER: $COUNTER/$TOTAL"
                                    echo " "
                                    COUNTER=$(( COUNTER + 1 ))
                                    python features.py \
                                    --dataset ${DATASET} \
                                    --train-shot ${SHOTS} \
                                    --clip-encoder ${ENCODER} \
                                    --image-layer-idx ${IMAGE_LAYER} \
                                    --text-layer-idx ${TEXT_LAYER} \
                                    --image-augmentation ${IMAGE_AUG} \
                                    --text-augmentation ${TEXT_AUG} \
                                    --image-views ${IMAGE_VIEW} \
                                    --seed ${SEED}
                                done
                            done
                        done
                    done
                done
            done
        done
    done
done