{
  "id": 415641,
  "title": "unet error ",
  "url": "/competitions/google-research-identify-contrails-reduce-global-warming/discussion/415641",
  "author_name": "Pranav Atote",
  "post_date": "2023-06-07T12:26:41.648000",
  "votes": 2,
  "comment_count": 0,
  "views": 0,
  "content": "<p>I  am new to kaggle and trying to make a unet in R in the competition notebook  sometimes it is working properly and sometime`get_unet_128 &lt;- function(input_shape = c(256, 256, 3),<br>\n                         num_classes = 1) {</p>\n<pre><code>inputs &lt;- layer_input(shape = input_shape)\n# \n\ndown1 &lt;- inputs %&gt;%\n    layer_conv_2d(filters = , kernel_size = c(, ), padding = ) %&gt;%\n    layer_batch_normalization() %&gt;%\n    layer_activation() %&gt;%\n    layer_conv_2d(filters = , kernel_size = c(, ), padding = ) %&gt;%\n    layer_batch_normalization() %&gt;%\n    layer_activation() \ndown1_pool &lt;- down1 %&gt;%\n    layer_max_pooling_2d(pool_size = c(, ), strides = c(, ))\n    # \n\ndown2 &lt;- down1_pool %&gt;%\n    layer_conv_2d(filters = , kernel_size = c(, ), padding = ) %&gt;%\n    layer_batch_normalization() %&gt;%\n    layer_activation() %&gt;%\n    layer_conv_2d(filters = , kernel_size = c(, ), padding = ) %&gt;%\n    layer_batch_normalization() %&gt;%\n    layer_activation() \ndown2_pool &lt;- down2 %&gt;%\n    layer_max_pooling_2d(pool_size = c(, ), strides = c(, ))\n    # \n\ndown3 &lt;- down2_pool %&gt;%\n    layer_conv_2d(filters = , kernel_size = c(, ), padding = ) %&gt;%\n    layer_batch_normalization() %&gt;%\n    layer_activation() %&gt;%\n    layer_conv_2d(filters = , kernel_size = c(, ), padding = ) %&gt;%\n    layer_batch_normalization() %&gt;%\n    layer_activation() \ndown3_pool &lt;- down3 %&gt;%\n    layer_max_pooling_2d(pool_size = c(, ), strides = c(, ))\n    # \n\ndown4 &lt;- down3_pool %&gt;%\n    layer_conv_2d(filters = , kernel_size = c(, ), padding = ) %&gt;%\n    layer_batch_normalization() %&gt;%\n    layer_activation() %&gt;%\n    layer_conv_2d(filters = , kernel_size = c(, ), padding = ) %&gt;%\n    layer_batch_normalization() %&gt;%\n    layer_activation() \ndown4_pool &lt;- down4 %&gt;%\n    layer_max_pooling_2d(pool_size = c(, ), strides = c(, ))\n    # \n\ncenter &lt;- down4_pool %&gt;%\n    layer_conv_2d(filters = , kernel_size = c(, ), padding = ) %&gt;%\n    layer_batch_normalization() %&gt;%\n    layer_activation() %&gt;%\n    layer_conv_2d(filters = , kernel_size = c(, ), padding = ) %&gt;%\n    layer_batch_normalization() %&gt;%\n    layer_activation() \n    # center\n\nup4 &lt;- center %&gt;%\n    layer_upsampling_2d( = c(, )) %&gt;%\n    layer_concatenate(down4, axis = ) %&gt;%\n    layer_conv_2d(filters = , kernel_size = c(, ), padding = ) %&gt;%\n    layer_batch_normalization() %&gt;%\n    layer_activation() %&gt;%\n    layer_conv_2d(filters = , kernel_size = c(, ), padding = ) %&gt;%\n    layer_batch_normalization() %&gt;%\n    layer_activation() %&gt;%\n    layer_conv_2d(filters = , kernel_size = c(, ), padding = ) %&gt;%\n    layer_batch_normalization() %&gt;%\n    layer_activation()\n    # \n\nup3 &lt;- up4 %&gt;%\n    layer_upsampling_2d( = c(, )) %&gt;%\n    layer_concatenate(down3, axis = ) %&gt;%\n    layer_conv_2d(filters = , kernel_size = c(, ), padding = ) %&gt;%\n    layer_batch_normalization() %&gt;%\n    layer_activation() %&gt;%\n    layer_conv_2d(filters = , kernel_size = c(, ), padding = ) %&gt;%\n    layer_batch_normalization() %&gt;%\n    layer_activation() %&gt;%\n    layer_conv_2d(filters = , kernel_size = c(, ), padding = ) %&gt;%\n    layer_batch_normalization() %&gt;%\n    layer_activation()\n    # \n\nup2 &lt;- up3 %&gt;%\n    layer_upsampling_2d( = c(, )) %&gt;%\n    layer_concatenate(down2, axis = ) %&gt;%\n    layer_conv_2d(filters = , kernel_size = c(, ), padding = ) %&gt;%\n    layer_batch_normalization() %&gt;%\n    layer_activation() %&gt;%\n    layer_conv_2d(filters = , kernel_size = c(, ), padding = ) %&gt;%\n    layer_batch_normalization() %&gt;%\n    layer_activation() %&gt;%\n    layer_conv_2d(filters = , kernel_size = c(, ), padding = ) %&gt;%\n    layer_batch_normalization() %&gt;%\n    layer_activation()\n    # \n\nup1 &lt;- up2 %&gt;%\n    layer_upsampling_2d( = c(, )) %&gt;%\n    layer_concatenate(down1, axis = ) %&gt;%\n    layer_conv_2d(filters = , kernel_size = c(, ), padding = ) %&gt;%\n    layer_batch_normalization() %&gt;%\n    layer_activation() %&gt;%\n    layer_conv_2d(filters = , kernel_size = c(, ), padding = ) %&gt;%\n    layer_batch_normalization() %&gt;%\n    layer_activation() %&gt;%\n    layer_conv_2d(filters = , kernel_size = c(, ), padding = ) %&gt;%\n    layer_batch_normalization() %&gt;%\n    layer_activation()\n    # \n\nclassify &lt;- layer_conv_2d(up1,\n                          filters = num_classes, \n                          kernel_size = c(, ),\n                          activation = )\n\n\nmodel &lt;- keras_model(\n    inputs = inputs,\n    outputs = classify\n)\n\nmodel %&gt;% compile(\n    optimizer = optimizer_rmsprop(lr = ),\n    loss = bce_dice_loss,\n    metrics = c(dice_coef)\n)\n\n(model)\n</code></pre>\n<p>}`s throwing error of trainable is none and it should be boolean.</p>",
  "messages": [
    {
      "id": 2291230,
      "postDate": "2023-06-07T12:26:41.647Z",
      "content": "<p>I  am new to kaggle and trying to make a unet in R in the competition notebook  sometimes it is working properly and sometime`get_unet_128 &lt;- function(input_shape = c(256, 256, 3),<br>\n                         num_classes = 1) {</p>\n<pre><code>inputs &lt;- layer_input(shape = input_shape)\n# \n\ndown1 &lt;- inputs %&gt;%\n    layer_conv_2d(filters = , kernel_size = c(, ), padding = ) %&gt;%\n    layer_batch_normalization() %&gt;%\n    layer_activation() %&gt;%\n    layer_conv_2d(filters = , kernel_size = c(, ), padding = ) %&gt;%\n    layer_batch_normalization() %&gt;%\n    layer_activation() \ndown1_pool &lt;- down1 %&gt;%\n    layer_max_pooling_2d(pool_size = c(, ), strides = c(, ))\n    # \n\ndown2 &lt;- down1_pool %&gt;%\n    layer_conv_2d(filters = , kernel_size = c(, ), padding = ) %&gt;%\n    layer_batch_normalization() %&gt;%\n    layer_activation() %&gt;%\n    layer_conv_2d(filters = , kernel_size = c(, ), padding = ) %&gt;%\n    layer_batch_normalization() %&gt;%\n    layer_activation() \ndown2_pool &lt;- down2 %&gt;%\n    layer_max_pooling_2d(pool_size = c(, ), strides = c(, ))\n    # \n\ndown3 &lt;- down2_pool %&gt;%\n    layer_conv_2d(filters = , kernel_size = c(, ), padding = ) %&gt;%\n    layer_batch_normalization() %&gt;%\n    layer_activation() %&gt;%\n    layer_conv_2d(filters = , kernel_size = c(, ), padding = ) %&gt;%\n    layer_batch_normalization() %&gt;%\n    layer_activation() \ndown3_pool &lt;- down3 %&gt;%\n    layer_max_pooling_2d(pool_size = c(, ), strides = c(, ))\n    # \n\ndown4 &lt;- down3_pool %&gt;%\n    layer_conv_2d(filters = , kernel_size = c(, ), padding = ) %&gt;%\n    layer_batch_normalization() %&gt;%\n    layer_activation() %&gt;%\n    layer_conv_2d(filters = , kernel_size = c(, ), padding = ) %&gt;%\n    layer_batch_normalization() %&gt;%\n    layer_activation() \ndown4_pool &lt;- down4 %&gt;%\n    layer_max_pooling_2d(pool_size = c(, ), strides = c(, ))\n    # \n\ncenter &lt;- down4_pool %&gt;%\n    layer_conv_2d(filters = , kernel_size = c(, ), padding = ) %&gt;%\n    layer_batch_normalization() %&gt;%\n    layer_activation() %&gt;%\n    layer_conv_2d(filters = , kernel_size = c(, ), padding = ) %&gt;%\n    layer_batch_normalization() %&gt;%\n    layer_activation() \n    # center\n\nup4 &lt;- center %&gt;%\n    layer_upsampling_2d( = c(, )) %&gt;%\n    layer_concatenate(down4, axis = ) %&gt;%\n    layer_conv_2d(filters = , kernel_size = c(, ), padding = ) %&gt;%\n    layer_batch_normalization() %&gt;%\n    layer_activation() %&gt;%\n    layer_conv_2d(filters = , kernel_size = c(, ), padding = ) %&gt;%\n    layer_batch_normalization() %&gt;%\n    layer_activation() %&gt;%\n    layer_conv_2d(filters = , kernel_size = c(, ), padding = ) %&gt;%\n    layer_batch_normalization() %&gt;%\n    layer_activation()\n    # \n\nup3 &lt;- up4 %&gt;%\n    layer_upsampling_2d( = c(, )) %&gt;%\n    layer_concatenate(down3, axis = ) %&gt;%\n    layer_conv_2d(filters = , kernel_size = c(, ), padding = ) %&gt;%\n    layer_batch_normalization() %&gt;%\n    layer_activation() %&gt;%\n    layer_conv_2d(filters = , kernel_size = c(, ), padding = ) %&gt;%\n    layer_batch_normalization() %&gt;%\n    layer_activation() %&gt;%\n    layer_conv_2d(filters = , kernel_size = c(, ), padding = ) %&gt;%\n    layer_batch_normalization() %&gt;%\n    layer_activation()\n    # \n\nup2 &lt;- up3 %&gt;%\n    layer_upsampling_2d( = c(, )) %&gt;%\n    layer_concatenate(down2, axis = ) %&gt;%\n    layer_conv_2d(filters = , kernel_size = c(, ), padding = ) %&gt;%\n    layer_batch_normalization() %&gt;%\n    layer_activation() %&gt;%\n    layer_conv_2d(filters = , kernel_size = c(, ), padding = ) %&gt;%\n    layer_batch_normalization() %&gt;%\n    layer_activation() %&gt;%\n    layer_conv_2d(filters = , kernel_size = c(, ), padding = ) %&gt;%\n    layer_batch_normalization() %&gt;%\n    layer_activation()\n    # \n\nup1 &lt;- up2 %&gt;%\n    layer_upsampling_2d( = c(, )) %&gt;%\n    layer_concatenate(down1, axis = ) %&gt;%\n    layer_conv_2d(filters = , kernel_size = c(, ), padding = ) %&gt;%\n    layer_batch_normalization() %&gt;%\n    layer_activation() %&gt;%\n    layer_conv_2d(filters = , kernel_size = c(, ), padding = ) %&gt;%\n    layer_batch_normalization() %&gt;%\n    layer_activation() %&gt;%\n    layer_conv_2d(filters = , kernel_size = c(, ), padding = ) %&gt;%\n    layer_batch_normalization() %&gt;%\n    layer_activation()\n    # \n\nclassify &lt;- layer_conv_2d(up1,\n                          filters = num_classes, \n                          kernel_size = c(, ),\n                          activation = )\n\n\nmodel &lt;- keras_model(\n    inputs = inputs,\n    outputs = classify\n)\n\nmodel %&gt;% compile(\n    optimizer = optimizer_rmsprop(lr = ),\n    loss = bce_dice_loss,\n    metrics = c(dice_coef)\n)\n\n(model)\n</code></pre>\n<p>}`s throwing error of trainable is none and it should be boolean.</p>",
      "rawMarkdown": "I  am new to kaggle and trying to make a unet in R in the competition notebook  sometimes it is working properly and sometime`get_unet_128 <- function(input_shape = c(256, 256, 3),\n                         num_classes = 1) {\n   \n    inputs <- layer_input(shape = input_shape)\n    # 128\n\n    down1 <- inputs %>%\n        layer_conv_2d(filters = 64, kernel_size = c(3, 3), padding = \"same\") %>%\n        layer_batch_normalization() %>%\n        layer_activation(\"relu\") %>%\n        layer_conv_2d(filters = 64, kernel_size = c(3, 3), padding = \"same\") %>%\n        layer_batch_normalization() %>%\n        layer_activation(\"relu\") \n    down1_pool <- down1 %>%\n        layer_max_pooling_2d(pool_size = c(2, 2), strides = c(2, 2))\n        # 64\n    \n    down2 <- down1_pool %>%\n        layer_conv_2d(filters = 128, kernel_size = c(3, 3), padding = \"same\") %>%\n        layer_batch_normalization() %>%\n        layer_activation(\"relu\") %>%\n        layer_conv_2d(filters = 128, kernel_size = c(3, 3), padding = \"same\") %>%\n        layer_batch_normalization() %>%\n        layer_activation(\"relu\") \n    down2_pool <- down2 %>%\n        layer_max_pooling_2d(pool_size = c(2, 2), strides = c(2, 2))\n        # 32\n\n    down3 <- down2_pool %>%\n        layer_conv_2d(filters = 256, kernel_size = c(3, 3), padding = \"same\") %>%\n        layer_batch_normalization() %>%\n        layer_activation(\"relu\") %>%\n        layer_conv_2d(filters = 256, kernel_size = c(3, 3), padding = \"same\") %>%\n        layer_batch_normalization() %>%\n        layer_activation(\"relu\") \n    down3_pool <- down3 %>%\n        layer_max_pooling_2d(pool_size = c(2, 2), strides = c(2, 2))\n        # 16\n\n    down4 <- down3_pool %>%\n        layer_conv_2d(filters = 512, kernel_size = c(3, 3), padding = \"same\") %>%\n        layer_batch_normalization() %>%\n        layer_activation(\"relu\") %>%\n        layer_conv_2d(filters = 512, kernel_size = c(3, 3), padding = \"same\") %>%\n        layer_batch_normalization() %>%\n        layer_activation(\"relu\") \n    down4_pool <- down4 %>%\n        layer_max_pooling_2d(pool_size = c(2, 2), strides = c(2, 2))\n        # 8\n    \n    center <- down4_pool %>%\n        layer_conv_2d(filters = 1024, kernel_size = c(3, 3), padding = \"same\") %>%\n        layer_batch_normalization() %>%\n        layer_activation(\"relu\") %>%\n        layer_conv_2d(filters = 1024, kernel_size = c(3, 3), padding = \"same\") %>%\n        layer_batch_normalization() %>%\n        layer_activation(\"relu\") \n        # center\n  \n    up4 <- center %>%\n        layer_upsampling_2d(size = c(2, 2)) %>%\n        layer_concatenate(down4, axis = 3) %>%\n        layer_conv_2d(filters = 512, kernel_size = c(3, 3), padding = \"same\") %>%\n        layer_batch_normalization() %>%\n        layer_activation(\"relu\") %>%\n        layer_conv_2d(filters = 512, kernel_size = c(3, 3), padding = \"same\") %>%\n        layer_batch_normalization() %>%\n        layer_activation(\"relu\") %>%\n        layer_conv_2d(filters = 512, kernel_size = c(3, 3), padding = \"same\") %>%\n        layer_batch_normalization() %>%\n        layer_activation(\"relu\")\n        # 16\n\n    up3 <- up4 %>%\n        layer_upsampling_2d(size = c(2, 2)) %>%\n        layer_concatenate(down3, axis = 3) %>%\n        layer_conv_2d(filters = 256, kernel_size = c(3, 3), padding = \"same\") %>%\n        layer_batch_normalization() %>%\n        layer_activation(\"relu\") %>%\n        layer_conv_2d(filters = 256, kernel_size = c(3, 3), padding = \"same\") %>%\n        layer_batch_normalization() %>%\n        layer_activation(\"relu\") %>%\n        layer_conv_2d(filters = 256, kernel_size = c(3, 3), padding = \"same\") %>%\n        layer_batch_normalization() %>%\n        layer_activation(\"relu\")\n        # 32\n\n    up2 <- up3 %>%\n        layer_upsampling_2d(size = c(2, 2)) %>%\n        layer_concatenate(down2, axis = 3) %>%\n        layer_conv_2d(filters = 128, kernel_size = c(3, 3), padding = \"same\") %>%\n        layer_batch_normalization() %>%\n        layer_activation(\"relu\") %>%\n        layer_conv_2d(filters = 128, kernel_size = c(3, 3), padding = \"same\") %>%\n        layer_batch_normalization() %>%\n        layer_activation(\"relu\") %>%\n        layer_conv_2d(filters = 128, kernel_size = c(3, 3), padding = \"same\") %>%\n        layer_batch_normalization() %>%\n        layer_activation(\"relu\")\n        # 64\n\n    up1 <- up2 %>%\n        layer_upsampling_2d(size = c(2, 2)) %>%\n        layer_concatenate(down1, axis = 3) %>%\n        layer_conv_2d(filters = 64, kernel_size = c(3, 3), padding = \"same\") %>%\n        layer_batch_normalization() %>%\n        layer_activation(\"relu\") %>%\n        layer_conv_2d(filters = 64, kernel_size = c(3, 3), padding = \"same\") %>%\n        layer_batch_normalization() %>%\n        layer_activation(\"relu\") %>%\n        layer_conv_2d(filters = 64, kernel_size = c(3, 3), padding = \"same\") %>%\n        layer_batch_normalization() %>%\n        layer_activation(\"relu\")\n        # 128\n\n    classify <- layer_conv_2d(up1,\n                              filters = num_classes, \n                              kernel_size = c(1, 1),\n                              activation = \"sigmoid\")\n\n\n    model <- keras_model(\n        inputs = inputs,\n        outputs = classify\n    )\n\n    model %>% compile(\n        optimizer = optimizer_rmsprop(lr = 0.0001),\n        loss = bce_dice_loss,\n        metrics = c(dice_coef)\n    )\n        \n    return(model)\n}`s throwing error of trainable is none and it should be boolean.",
      "votes": 1
    }
  ],
  "comments": [],
  "raw_markdown_by_id": {
    "2291230": "I  am new to kaggle and trying to make a unet in R in the competition notebook  sometimes it is working properly and sometime`get_unet_128 <- function(input_shape = c(256, 256, 3),\n                         num_classes = 1) {\n   \n    inputs <- layer_input(shape = input_shape)\n    # 128\n\n    down1 <- inputs %>%\n        layer_conv_2d(filters = 64, kernel_size = c(3, 3), padding = \"same\") %>%\n        layer_batch_normalization() %>%\n        layer_activation(\"relu\") %>%\n        layer_conv_2d(filters = 64, kernel_size = c(3, 3), padding = \"same\") %>%\n        layer_batch_normalization() %>%\n        layer_activation(\"relu\") \n    down1_pool <- down1 %>%\n        layer_max_pooling_2d(pool_size = c(2, 2), strides = c(2, 2))\n        # 64\n    \n    down2 <- down1_pool %>%\n        layer_conv_2d(filters = 128, kernel_size = c(3, 3), padding = \"same\") %>%\n        layer_batch_normalization() %>%\n        layer_activation(\"relu\") %>%\n        layer_conv_2d(filters = 128, kernel_size = c(3, 3), padding = \"same\") %>%\n        layer_batch_normalization() %>%\n        layer_activation(\"relu\") \n    down2_pool <- down2 %>%\n        layer_max_pooling_2d(pool_size = c(2, 2), strides = c(2, 2))\n        # 32\n\n    down3 <- down2_pool %>%\n        layer_conv_2d(filters = 256, kernel_size = c(3, 3), padding = \"same\") %>%\n        layer_batch_normalization() %>%\n        layer_activation(\"relu\") %>%\n        layer_conv_2d(filters = 256, kernel_size = c(3, 3), padding = \"same\") %>%\n        layer_batch_normalization() %>%\n        layer_activation(\"relu\") \n    down3_pool <- down3 %>%\n        layer_max_pooling_2d(pool_size = c(2, 2), strides = c(2, 2))\n        # 16\n\n    down4 <- down3_pool %>%\n        layer_conv_2d(filters = 512, kernel_size = c(3, 3), padding = \"same\") %>%\n        layer_batch_normalization() %>%\n        layer_activation(\"relu\") %>%\n        layer_conv_2d(filters = 512, kernel_size = c(3, 3), padding = \"same\") %>%\n        layer_batch_normalization() %>%\n        layer_activation(\"relu\") \n    down4_pool <- down4 %>%\n        layer_max_pooling_2d(pool_size = c(2, 2), strides = c(2, 2))\n        # 8\n    \n    center <- down4_pool %>%\n        layer_conv_2d(filters = 1024, kernel_size = c(3, 3), padding = \"same\") %>%\n        layer_batch_normalization() %>%\n        layer_activation(\"relu\") %>%\n        layer_conv_2d(filters = 1024, kernel_size = c(3, 3), padding = \"same\") %>%\n        layer_batch_normalization() %>%\n        layer_activation(\"relu\") \n        # center\n  \n    up4 <- center %>%\n        layer_upsampling_2d(size = c(2, 2)) %>%\n        layer_concatenate(down4, axis = 3) %>%\n        layer_conv_2d(filters = 512, kernel_size = c(3, 3), padding = \"same\") %>%\n        layer_batch_normalization() %>%\n        layer_activation(\"relu\") %>%\n        layer_conv_2d(filters = 512, kernel_size = c(3, 3), padding = \"same\") %>%\n        layer_batch_normalization() %>%\n        layer_activation(\"relu\") %>%\n        layer_conv_2d(filters = 512, kernel_size = c(3, 3), padding = \"same\") %>%\n        layer_batch_normalization() %>%\n        layer_activation(\"relu\")\n        # 16\n\n    up3 <- up4 %>%\n        layer_upsampling_2d(size = c(2, 2)) %>%\n        layer_concatenate(down3, axis = 3) %>%\n        layer_conv_2d(filters = 256, kernel_size = c(3, 3), padding = \"same\") %>%\n        layer_batch_normalization() %>%\n        layer_activation(\"relu\") %>%\n        layer_conv_2d(filters = 256, kernel_size = c(3, 3), padding = \"same\") %>%\n        layer_batch_normalization() %>%\n        layer_activation(\"relu\") %>%\n        layer_conv_2d(filters = 256, kernel_size = c(3, 3), padding = \"same\") %>%\n        layer_batch_normalization() %>%\n        layer_activation(\"relu\")\n        # 32\n\n    up2 <- up3 %>%\n        layer_upsampling_2d(size = c(2, 2)) %>%\n        layer_concatenate(down2, axis = 3) %>%\n        layer_conv_2d(filters = 128, kernel_size = c(3, 3), padding = \"same\") %>%\n        layer_batch_normalization() %>%\n        layer_activation(\"relu\") %>%\n        layer_conv_2d(filters = 128, kernel_size = c(3, 3), padding = \"same\") %>%\n        layer_batch_normalization() %>%\n        layer_activation(\"relu\") %>%\n        layer_conv_2d(filters = 128, kernel_size = c(3, 3), padding = \"same\") %>%\n        layer_batch_normalization() %>%\n        layer_activation(\"relu\")\n        # 64\n\n    up1 <- up2 %>%\n        layer_upsampling_2d(size = c(2, 2)) %>%\n        layer_concatenate(down1, axis = 3) %>%\n        layer_conv_2d(filters = 64, kernel_size = c(3, 3), padding = \"same\") %>%\n        layer_batch_normalization() %>%\n        layer_activation(\"relu\") %>%\n        layer_conv_2d(filters = 64, kernel_size = c(3, 3), padding = \"same\") %>%\n        layer_batch_normalization() %>%\n        layer_activation(\"relu\") %>%\n        layer_conv_2d(filters = 64, kernel_size = c(3, 3), padding = \"same\") %>%\n        layer_batch_normalization() %>%\n        layer_activation(\"relu\")\n        # 128\n\n    classify <- layer_conv_2d(up1,\n                              filters = num_classes, \n                              kernel_size = c(1, 1),\n                              activation = \"sigmoid\")\n\n\n    model <- keras_model(\n        inputs = inputs,\n        outputs = classify\n    )\n\n    model %>% compile(\n        optimizer = optimizer_rmsprop(lr = 0.0001),\n        loss = bce_dice_loss,\n        metrics = c(dice_coef)\n    )\n        \n    return(model)\n}`s throwing error of trainable is none and it should be boolean."
  }
}