{"id":28629777,"url":"https://github.com/codebox/convnet-designer","last_synced_at":"2025-06-12T12:13:34.644Z","repository":{"id":140514510,"uuid":"55442435","full_name":"codebox/convnet-designer","owner":"codebox","description":"A utility for designing Convolutional Neural Networks","archived":false,"fork":false,"pushed_at":"2019-03-06T08:33:40.000Z","size":103,"stargazers_count":27,"open_issues_count":2,"forks_count":10,"subscribers_count":5,"default_branch":"master","last_synced_at":"2023-10-20T18:59:17.667Z","etag":null,"topics":["convolutional-neural-networks","tensorflow"],"latest_commit_sha":null,"homepage":"https://codebox.net/pages/tensorflow-convnet-designer","language":"JavaScript","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"mit","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/codebox.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":"LICENSE","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null}},"created_at":"2016-04-04T20:27:37.000Z","updated_at":"2023-10-20T18:59:23.618Z","dependencies_parsed_at":null,"dependency_job_id":"f35e926f-2bca-4840-9e3a-5d9a11025d43","html_url":"https://github.com/codebox/convnet-designer","commit_stats":null,"previous_names":[],"tags_count":0,"template":null,"template_full_name":null,"purl":"pkg:github/codebox/convnet-designer","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/codebox%2Fconvnet-designer","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/codebox%2Fconvnet-designer/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/codebox%2Fconvnet-designer/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/codebox%2Fconvnet-designer/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/codebox","download_url":"https://codeload.github.com/codebox/convnet-designer/tar.gz/refs/heads/master","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/codebox%2Fconvnet-designer/sbom","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":259462578,"owners_count":22861514,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2022-07-04T15:15:14.044Z","host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["convolutional-neural-networks","tensorflow"],"created_at":"2025-06-12T12:13:33.921Z","updated_at":"2025-06-12T12:13:34.617Z","avatar_url":"https://github.com/codebox.png","language":"JavaScript","readme":"#  Convolutional Neural Network Designer\n\nThis utility is intended to help anyone who needs to design a Convolutional Neural Network, and may be particularly useful if using [Google's Tensorflow library](https://www.tensorflow.org/).\n\nThe utility is [available to use online](https://codebox.net/convnetdesigner/main.html), or can be downloaded from GitHub and used offline by opening the main.html file in a web browser. \n\nThe utility performs the following functions:\n* Visualises the network, showing the relative layer sizes and helping to ensure that the overal 'shape' of the network is correct:\n    \n\u003cimg alt=\"Convnet Designer visualising a network\" src=\"https://codebox.net/assets/images/convnet_designer2.png\" style=\"display: block; margin: 0 auto; border: 1px solid grey\"/\u003e\n\n* Computes the total number of parameters required by the network, allowing the computational complexity of different designs to be compared\n* Estimates the amount of memory required by the network\n* Validates patch size and stride values for convolutional layers, automatically applying zero-padding if required\n\n\u003cimg alt=\"Convnet Designer layer inputs\" src=\"https://codebox.net/assets/images/convnet_designer1.png\" style=\"display: block; margin: 0 auto; border: 1px solid grey\"/\u003e\n\n* Validates pool size and stride values for pooling layers\n* Generates the Tensorflow code required to construct the network (example code below)\n\n\u003cpre\u003e\n    x0 = tf.placeholder(tf.float32, shape=[None, 49152])\n    \n    x1 = tf.reshape(x0, [-1,128,128,3])\n    x2 = tf.Variable(tf.truncated_normal([3, 3, 3, 32], stddev=0.1))\n    x3 = tf.nn.conv2d(x1, x2, strides=[1, 1, 1, 1], padding='SAME')\n    \n    x4 = tf.nn.max_pool(x3, ksize=[1, 2, 2, 1], strides=[1, 2, 2, 1], padding='SAME')\n    \n    x5 = tf.Variable(tf.constant(0.1, shape=[32]))\n    x6 = tf.nn.relu(x4 + x5)\n    \n    x7 = tf.reshape(x6, [-1,64,64,32])\n    x8 = tf.Variable(tf.truncated_normal([3, 3, 32, 64], stddev=0.1))\n    x9 = tf.nn.conv2d(x7, x8, strides=[1, 1, 1, 1], padding='SAME')\n    \n    x10 = tf.nn.max_pool(x9, ksize=[1, 2, 2, 1], strides=[1, 2, 2, 1], padding='SAME')\n    \n    x11 = tf.Variable(tf.constant(0.1, shape=[64]))\n    x12 = tf.nn.relu(x10 + x11)\n    \n    x13 = tf.Variable(tf.truncated_normal([65536, 1024], stddev=0.1))\n    x14 = tf.reshape(x12, [-1, 65536])\n    x15 = tf.matmul(x14, x13)\n    \n    x16 = tf.Variable(tf.truncated_normal([1024, 2], stddev=0.1))\n    x17 = tf.reshape(x15, [-1, 1024])\n    y_conv = tf.matmul(x17, x16)\n\u003c/pre\u003e\n","funding_links":[],"categories":[],"sub_categories":[],"project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fcodebox%2Fconvnet-designer","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fcodebox%2Fconvnet-designer","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fcodebox%2Fconvnet-designer/lists"}