dandelin / Dynamic Memory Networks Plus Pytorch
Implementation of Dynamic memory networks plus in Pytorch
Stars: ✭ 123
Programming Languages
python
139335 projects - #7 most used programming language
Projects that are alternatives of or similar to Dynamic Memory Networks Plus Pytorch
Bert language understanding
Pre-training of Deep Bidirectional Transformers for Language Understanding: pre-train TextCNN
Stars: ✭ 933 (+658.54%)
Mutual labels: question-answering, language-model
backprop
Backprop makes it simple to use, finetune, and deploy state-of-the-art ML models.
Stars: ✭ 229 (+86.18%)
Mutual labels: question-answering, language-model
Spago
Self-contained Machine Learning and Natural Language Processing library in Go
Stars: ✭ 854 (+594.31%)
Mutual labels: question-answering, language-model
Nlp chinese corpus
大规模中文自然语言处理语料 Large Scale Chinese Corpus for NLP
Stars: ✭ 6,656 (+5311.38%)
Mutual labels: question-answering, language-model
Haystack
🔍 Haystack is an open source NLP framework that leverages Transformer models. It enables developers to implement production-ready neural search, question answering, semantic document search and summarization for a wide range of applications.
Stars: ✭ 3,409 (+2671.54%)
Mutual labels: question-answering, language-model
Flexneuart
Flexible classic and NeurAl Retrieval Toolkit
Stars: ✭ 99 (-19.51%)
Mutual labels: question-answering
Tableqa
AI Tool for querying natural language on tabular data.
Stars: ✭ 109 (-11.38%)
Mutual labels: question-answering
Vqa Tensorflow
Tensorflow Implementation of Deeper LSTM+ normalized CNN for Visual Question Answering
Stars: ✭ 98 (-20.33%)
Mutual labels: question-answering
Sentence Similarity
PyTorch implementations of various deep learning models for paraphrase detection, semantic similarity, and textual entailment
Stars: ✭ 96 (-21.95%)
Mutual labels: question-answering
Lingo
package lingo provides the data structures and algorithms required for natural language processing
Stars: ✭ 113 (-8.13%)
Mutual labels: language-model
Reading Comprehension Question Answering Papers
Survey on Machine Reading Comprehension
Stars: ✭ 101 (-17.89%)
Mutual labels: question-answering
Bi Att Flow
Bi-directional Attention Flow (BiDAF) network is a multi-stage hierarchical process that represents context at different levels of granularity and uses a bi-directional attention flow mechanism to achieve a query-aware context representation without early summarization.
Stars: ✭ 1,472 (+1096.75%)
Mutual labels: question-answering
Neuronblocks
NLP DNN Toolkit - Building Your NLP DNN Models Like Playing Lego
Stars: ✭ 1,356 (+1002.44%)
Mutual labels: question-answering
Dynamic Coattention Network Plus
Dynamic Coattention Network Plus (DCN+) TensorFlow implementation. Question answering using Deep NLP.
Stars: ✭ 117 (-4.88%)
Mutual labels: question-answering
Happy Transformer
A package built on top of Hugging Face's transformer library that makes it easy to utilize state-of-the-art NLP models
Stars: ✭ 97 (-21.14%)
Mutual labels: question-answering
Easy Bert
A Dead Simple BERT API for Python and Java (https://github.com/google-research/bert)
Stars: ✭ 106 (-13.82%)
Mutual labels: language-model
Keras Gpt 2
Load GPT-2 checkpoint and generate texts
Stars: ✭ 113 (-8.13%)
Mutual labels: language-model
Ama
[[I'm slow at replying these days, but I hope to get back to answering questions eventually]] Ask me anything!
Stars: ✭ 102 (-17.07%)
Mutual labels: question-answering
Dynamic-memory-networks-plus-Pytorch
DMN+ implementation in Pytorch for question answering on the bAbI 10k dataset.
Contents
| file | description |
|---|---|
babi_loader.py |
declaration of bAbI Pytorch Dataset class |
babi_main.py |
contains DMN+ model and training code |
fetch_data.sh |
shell script to fetch bAbI tasks (from DMNs in Theano) |
Usage
Install Pytorch v0.1.12 and Python 3.6.x (for Literal String Interpolation)
Run the included shell script to fetch the data
chmod +x fetch_data.sh
./fetch_data.sh
Run the main python code
python babi_main.py
Benchmarks
Low accuracies compared to Xiong et al's are may due to different weight decay setting or the model's instability.
On some tasks, the accuracy was not stable across multiple runs. This was particularly problematic on QA3, QA17, and QA18. To solve this, we repeated training 10 times using random initializations and evaluated the model that achieved the lowest validation set loss.
You can find pretrained models here
| Task ID | This Repo | Xiong et al |
|---|---|---|
| 1 | 100% | 100% |
| 2 | 96.8% | 99.7% |
| 3 | 89.2% | 98.9% |
| 4 | 100% | 100% |
| 5 | 99.5% | 99.5% |
| 6 | 100% | 100% |
| 7 | 97.8% | 97.6% |
| 8 | 100% | 100% |
| 9 | 100% | 100% |
| 10 | 100% | 100% |
| 11 | 100% | 100% |
| 12 | 100% | 100% |
| 13 | 100% | 100% |
| 14 | 99% | 99.8% |
| 15 | 100% | 100% |
| 16 | 51.6% | 54.7% |
| 17 | 86.4% | 95.8% |
| 18 | 97.9% | 97.9% |
| 19 | 99.7% | 100% |
| 20 | 100% | 100% |
Note that the project description data, including the texts, logos, images, and/or trademarks,
for each open source project belongs to its rightful owner.
If you wish to add or remove any projects, please contact us at [email protected].