addons

addons copied to clipboard

Multi-class Precision and Recall

5

**Describe the feature and the current behavior/state.** Please add multi-class precision and recall metrics, much like that in `sklearn.metrics`. Currently, `tf.metrics.Precision` and `tf.metrics.Recall` only support binary labels. `sklearn.metrics` supports averages...

jraman

No gradients to transforms in tfa.image.transform

1

**Describe the feature and the current behavior/state.** back prop the gradient to transforms. **Relevant information** - Are you willing to contribute it (yes/no): no - Are you willing to maintain...

fengyang0317

EmbeddingBag for ragged indices

5

**Describe the feature and the current behavior/state.** Currently EmbeddingBag and underlying op works only with dense tensors. But a lot of nlp ops/tasks in modern require to use RaggedTensors. Converting...

shkarupa-alex

Hello, I tried to use F1Score metrics instead of accuracy during training. I get an error with incompatible shapes because I used sparse_categorical_crossentropy and my labels are sparsely defined. Please...

markub3327

Loss calculated incorrectly in networks_seq2seq_nmt.ipynb

4

**System information** - OS Platform and Distribution (e.g., Linux Ubuntu 16.04): N/A - TensorFlow version and how it was installed (source or binary): N/A - TensorFlow-Addons version and how it...

martingoodson

Proposed feature: Multi-Head Attention with O(sqrt(N)) memory

3

The recent paper by Rabe, Staats, et. al on the Google Research team ["Self-attention Does Not Need O(n2) Memory"](https://arxiv.org/abs/2112.05682) provides a numerically stable implementation of multi-headed attention that requires O(√n)...

singularperturbation

Add Mixout module

7

It's an issue switched from tf repo ([here](https://github.com/tensorflow/tensorflow/issues/40983)) and here is [the pending pr](https://github.com/tensorflow/tensorflow/pull/40990) I've sent there. According to the reviewer's suggestion, I probably should add the module to here...

crystina-z

add CRF tutorial

10

# Description Brief Description of the PR: add a tutorial for CRF Layer **To Reviewers**: Here is the candidate feature list for this tutorial: * [Done] Replace faked training data...

howl-anderson

NaN Loss bug and a step-by-step code analysis

2

Following on problems I had while training with Triplet Loss, I went through the backend code and gathered some personal / visual explanations. Feel free to use my resources. https://tudorvladstefan.medium.com/de-mystifying-tensorflows-triplet-loss-deca00ca9479

tdrvlad

dev_container

10

**System information** - OS Platform and Distribution (e.g., Linux Ubuntu 16.04): - TensorFlow version and how it was installed (source or binary): - TensorFlow-Addons version and how it was installed...

fsx950223