searchspace.json

{
    "handwrite/__line6/choice": {
        "_type": "choice",
        "_value": [
            "foo",
            "bar"
        ]
    },
    "handwrite/conv_size/choice": {
        "_type": "choice",
        "_value": [
            2,
            3,
            5,
            7
        ]
    },
    "handwrite/abc/choice": {
        "_type": "choice",
        "_value": [
            "2",
            3,
            "(5 * 6)",
            7
        ]
    },
    "handwrite/__line9/function_choice": {
        "_type": "choice",
        "_value": [
            "max_pool",
            "h_conv1",
            "avg_pool"
        ]
    },
    "handwrite/max_pool/function_choice": {
        "_type": "choice",
        "_value": [
            "max_pool(h_conv1)",
            "avg_pool(h_conv2, h_conv3)"
        ]
    },
    "handwrite/max_poo/function_choice": {
        "_type": "choice",
        "_value": [
            "max_poo(h_conv1)",
            "(2 * 3 + 4)",
            "(lambda x: 1 + x)"
        ]
    },
    "handwrite/__line19/qlognormal": {
        "_type": "qlognormal",
        "_value": [
            1.2,
            3,
            4.5
        ]
    },
    "mnist/self.conv_size/choice": {
        "_type": "choice",
        "_value": [
            2,
            3,
            5,
            7
        ]
    },
    "mnist/self.hidden_size/choice": {
        "_type": "choice",
        "_value": [
            124,
            512,
            1024
        ]
    },
    "mnist/self.learning_rate/randint": {
        "_type": "randint",
        "_value": [
            2,
            3,
            5
        ]
    },
    "mnist/tf.nn.relu/function_choice": {
        "_type": "choice",
        "_value": [
            "tf.nn.relu(conv2d(x_image, W_conv1) + b_conv1)",
            "tf.nn.sigmoid(conv2d(x_image, W_conv1) + b_conv1)",
            "tf.nn.tanh(conv2d(x_image, W_conv1) + b_conv1)"
        ]
    },
    "mnist/max_pool/function_choice": {
        "_type": "choice",
        "_value": [
            "max_pool(h_conv1, self.pool_size)",
            "avg_pool(h_conv1, self.pool_size)"
        ]
    },
    "mnist/batch_size/choice": {
        "_type": "choice",
        "_value": [
            50,
            250,
            500
        ]
    },
    "mnist/dropout_rate/choice": {
        "_type": "choice",
        "_value": [
            1,
            5
        ]
    },
    "dir.simple/conv_size/choice": {
        "_type": "choice",
        "_value": [
            2,
            3,
            5,
            7
        ]
    },
    "dir.simple/abc/choice": {
        "_type": "choice",
        "_value": [
            "2",
            3,
            "(5 * 6)",
            "{(1): 2, '3': 4}",
            "[1, 2, 3]"
        ]
    },
    "dir.simple/max_pool/function_choice": {
        "_type": "choice",
        "_value": [
            "max_pool(h_conv1)",
            "avg_pool(h_conv2, h_conv3)"
        ]
    },
    "dir.simple/max_poo/function_choice": {
        "_type": "choice",
        "_value": [
            "max_poo(h_conv1)",
            "(2 * 3 + 4)",
            "(lambda x: 1 + x)"
        ]
    }
}