tatp22/multidim-positional-encoding
An implementation of 1D, 2D, and 3D positional encoding in Pytorch and TensorFlow
This tool helps machine learning engineers incorporate spatial information into their deep learning models. It takes numerical data representing sequences, images, or 3D volumes and adds specific positional markers to these inputs. This allows models to understand the order or location of data points, which is critical for tasks like processing video, medical scans, or other structured data. It's used by machine learning engineers building models with PyTorch or TensorFlow.
615 stars. No commits in the last 6 months.
Use this if your deep learning models need to understand the relative position of elements within 1D sequences, 2D images, or 3D volumes to improve their performance.
Not ideal if you are working with unstructured data where spatial relationships are not relevant, or if you prefer a different framework than PyTorch or TensorFlow.
Stars
615
Forks
36
Language
Python
License
MIT
Category
Last pushed
Oct 23, 2024
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/tatp22/multidim-positional-encoding"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
philipperemy/keras-attention
Keras Attention Layer (Luong and Bahdanau scores).
tatp22/linformer-pytorch
My take on a practical implementation of Linformer for Pytorch.
datalogue/keras-attention
Visualizing RNNs using the attention mechanism
ematvey/hierarchical-attention-networks
Document classification with Hierarchical Attention Networks in TensorFlow. WARNING: project is...
thushv89/attention_keras
Keras Layer implementation of Attention for Sequential models