René's URL Explorer Experiment

{"@context":"https://schema.org","@type":"DiscussionForumPosting","headline":"Automatic Differentiation and Gradients tf.GradientTape()","articleBody":"\u003cem\u003ePlease make sure that this is a feature request. As per our [GitHub Policy](https://github.com/tensorflow/tensorflow/blob/master/ISSUES.md), we only address code/doc bugs, performance issues, feature requests and build/installation issues on GitHub. tag:feature_template\u003c/em\u003e\r\n\r\n\r\n**System information**\r\n- TensorFlow version (you are using): 2.3.1\r\n- Are you willing to contribute it (Yes/No): Yes, when able and available\r\n\r\n\r\n**Describe the feature and the current behavior/state.**\r\nTensorFlow provides the `tf.GradientTape` API to differentiate automatically, TensorFlow needs to remember what operations happen in what order during the forward pass. Then, during the backward pass, TensorFlow traverses this list of operations in reverse order to compute gradients.\r\n\r\nDetails about this feature can be found in the official TensorFlow documentation for [Gradient Tapes](https://www.tensorflow.org/guide/autodiff)\r\n\r\n**Will this change the current api? How?**\r\nYes, I think it will add a new feature to `tensorflow-core` module \r\n\r\n**Who will benefit with this feature?**\r\nAnyone that requires a very low-level control over training and evaluation of a deep learning model and everyone who is already familiar with TF/Keras.\r\n\r\n**Any Other info.**\r\n`tf.GradientTape` API is needed when writing a training loop from scratch as described [here](https://www.tensorflow.org/guide/keras/writing_a_training_loop_from_scratch/)\r\n","author":{"url":"https://github.com/danilojsl","@type":"Person","name":"danilojsl"},"datePublished":"2021-04-11T09:24:23.000Z","interactionStatistic":{"@type":"InteractionCounter","interactionType":"https://schema.org/CommentAction","userInteractionCount":2},"url":"https://github.com/285/java/issues/285"}