diffmem

Memory models

Differentiable memory schemes for neural networks

Reference implementations of neural networks with differentiable memory mechanisms (NTM, Stack RNN, etc.)

GitHub

220 stars
13 watching
29 forks
Language: Python
last commit: over 10 years ago

Related projects:

RepositoryDescriptionStars
loudinthecloud/pytorch-ntmAn implementation of Neural Turing Machines in PyTorch592
vonfeng/deepmoveA PyTorch-based implementation of a predictive model for human mobility patterns using attention mechanisms and recurrent neural networks.149
ypxie/pytorch-neucomAn implementation of the Differentiable Neural Computer architecture in PyTorch94
l0sg/relational-rnn-pytorchAn implementation of DeepMind's Relational Recurrent Neural Networks (Santoro et al. 2018) in PyTorch for word language modeling245
mhlee0903/multi_channels_pinnInvestigating neural networks for drug discovery using multiple chemical descriptors.3
nttcslab-nlp/doc_lmThis repository contains source files and training scripts for language models.12
jingweiz/pytorch-dncAn implementation of Neural Turing Machine and Differentiable Neural Computer architectures using PyTorch and Visdom for deep learning tasks.278
elanmart/psmmAn implementation of a neural network model for character-level language modeling.50
tdeboissiere/deeplearningimplementationsA collection of implementations of recent deep learning papers in Python1,814
rwightman/pytorch-dpn-pretrainedA PyTorch implementation of Dual Path Networks (DPNs) with pre-trained weights and conversion scripts.206
michael-wzhu/shennong-tcm-llmDevelops and deploys a large language model for Chinese traditional medicine applications316
ahmedfgad/numpyannAn implementation of artificial neural networks using NumPy98
oyam/pytorch-dpnsPyTorch implementation of a deep learning model for image segmentation90
j4mie/micromodelsA library that provides declarative dictionary-based model classes for building and serializing data structures in Python.103
vlgiitr/dmn-plusA PyTorch implementation of an improved question answering architecture with dynamic memory networks and attention mechanisms64