{"id":19464371,"url":"https://github.com/aryaaftab/dropconnect-tensorflow","last_synced_at":"2025-04-25T09:31:27.357Z","repository":{"id":57424211,"uuid":"400912006","full_name":"AryaAftab/dropconnect-tensorflow","owner":"AryaAftab","description":"An implementation of DropConnect Layer (Dense, Conv2D, and Wrapper(for all TensorFlow Layers)) in Tensorflow 2","archived":false,"fork":false,"pushed_at":"2021-09-12T04:24:54.000Z","size":21,"stargazers_count":5,"open_issues_count":0,"forks_count":2,"subscribers_count":1,"default_branch":"master","last_synced_at":"2025-04-02T00:30:06.342Z","etag":null,"topics":["deep-learning","dropconnect","dropconnect-lstm","tensorflow2"],"latest_commit_sha":null,"homepage":"","language":"Python","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/AryaAftab.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}},"created_at":"2021-08-28T23:45:22.000Z","updated_at":"2022-12-16T06:27:07.000Z","dependencies_parsed_at":"2022-09-14T08:41:41.284Z","dependency_job_id":null,"html_url":"https://github.com/AryaAftab/dropconnect-tensorflow","commit_stats":null,"previous_names":[],"tags_count":0,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/AryaAftab%2Fdropconnect-tensorflow","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/AryaAftab%2Fdropconnect-tensorflow/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/AryaAftab%2Fdropconnect-tensorflow/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/AryaAftab%2Fdropconnect-tensorflow/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/AryaAftab","download_url":"https://codeload.github.com/AryaAftab/dropconnect-tensorflow/tar.gz/refs/heads/master","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":250790062,"owners_count":21487739,"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":["deep-learning","dropconnect","dropconnect-lstm","tensorflow2"],"created_at":"2024-11-10T18:14:43.942Z","updated_at":"2025-04-25T09:31:27.069Z","avatar_url":"https://github.com/AryaAftab.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"## Drop Connect - Tensorflow\nAn implementation of \u003ca href=\"http://proceedings.mlr.press/v28/wan13.html\"\u003eDrop-Connect Layer\u003c/a\u003e \nin tensorflow 2.x. \nImplementation of layers of Dense, Conv2D, and Wrapper(for all TensorFlow Layers) has been done.\n\n## Demo\n[![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/AryaAftab/dropconnect-tensorflow/blob/master/demo/dropconnect_tensorflow_demo.ipynb)\n## Install\n\n```bash\n$ pip install dropconnect-tensorflow\n```\n\n## Usage\n\n### Fully-Connected Network\n```python\nimport tensorflow as tf\nfrom tensorflow.keras.layers import Dense, Input\nfrom dropconnect_tensorflow import DropConnectDense\n\n# Create Fully-Connected Network\nX = tf.keras.layers.Input(shape=(784,))\nx = DropConnectDense(units=128, prob=0.2, activation=\"relu\", use_bias=True)(X)\nx = DropConnectDense(units=64, prob=0.5, activation=\"relu\", use_bias=True)(x)\ny = Dense(10, activation=\"softmax\")(x)\n\nmodel = tf.keras.models.Model(X, y)\n\n\n# Hyperparameters\nbatch_size=64\nepochs=20\n\n# Compile the model\nmodel.compile(\n    optimizer=tf.keras.optimizers.Adam(0.0001),  # Utilize optimizer\n    loss=tf.keras.losses.SparseCategoricalCrossentropy(),\n    metrics=['accuracy'])\n\n# Train the network\nhistory = model.fit(\n    x_train,\n    y_train,\n    batch_size=batch_size,\n    validation_split=0.1,\n    epochs=epochs)\n```\n\n### Convolution Network\n```python\nimport tensorflow as tf\nfrom tensorflow.keras.layers import Dense, Conv2D, Input, ReLU, BatchNormalization, Flatten, MaxPool2D\nfrom dropconnect_tensorflow import DropConnectConv2D, DropConnectDense\n\n# Create Convolution Network\nX = tf.keras.layers.Input(shape=(28, 28, 1))\nx = DropConnectConv2D(filters=64, kernel_size=3, strides=(1, 1), padding='valid', prob=0.1)(X)\nx = BatchNormalization()(x)\nx = ReLU()(x)\nx = MaxPool2D((2,2))(x)\nx = DropConnectConv2D(filters=128, kernel_size=3, strides=(1, 1), padding='valid', prob=0.1)(x)\nx = BatchNormalization()(x)\nx = ReLU()(x)\nx = MaxPool2D((2,2))(x)\n\nx = Flatten()(x)\nx = DropConnectDense(units=64, prob=0.3, activation=\"relu\", use_bias=True)(x)\ny = Dense(10, activation=\"softmax\")(x)\n\nmodel = tf.keras.models.Model(X, y)\n\n\n# Hyperparameters\nbatch_size=64\nepochs=20\n\n# Compile the model\nmodel.compile(\n    optimizer=tf.keras.optimizers.Adam(0.0001),  # Utilize optimizer\n    loss=tf.keras.losses.SparseCategoricalCrossentropy(),\n    metrics=['accuracy'])\n\n# Train the network\nhistory = model.fit(\n    x_train,\n    y_train,\n    batch_size=batch_size,\n    validation_split=0.1,\n    epochs=epochs)\n```\n\n### Wrapper(GRU, LSTM, Dense, Con2D, Conv1D, ...) Network\n```python\nimport tensorflow as tf\nfrom tensorflow.keras.layers import Dense, Input, LSTM\nfrom dropconnect_tensorflow import DropConnect\n\n# Create LSTM Network\nX = tf.keras.layers.Input(shape=(28,28))\n\nx = DropConnect(LSTM(128, return_sequences=True), prob=0.5)(X)\nx = DropConnect(LSTM(128), prob=0.5)(X)\ny = Dense(10, activation=\"softmax\")(x)\n\nmodel = tf.keras.models.Model(X, y)\n\n\n# Hyperparameters\nbatch_size=64\nepochs=20\n\n# Compile the model\nmodel.compile(\n    optimizer=tf.keras.optimizers.Adam(0.0001),  # Utilize optimizer\n    loss=tf.keras.losses.SparseCategoricalCrossentropy(),\n    metrics=['accuracy'])\n\n# Train the network\nhistory = model.fit(\n    x_train,\n    y_train,\n    batch_size=batch_size,\n    validation_split=0.1,\n    epochs=epochs)\n```\n\n\n## Citations\n\n```bibtex\n@inproceedings{wan2013regularization,\n  title={Regularization of neural networks using dropconnect},\n  author={Wan, Li and Zeiler, Matthew and Zhang, Sixin and Le Cun, Yann and Fergus, Rob},\n  booktitle={International conference on machine learning},\n  pages={1058--1066},\n  year={2013},\n  organization={PMLR}\n}\n```\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Faryaaftab%2Fdropconnect-tensorflow","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Faryaaftab%2Fdropconnect-tensorflow","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Faryaaftab%2Fdropconnect-tensorflow/lists"}