kimhc6028 / Pathnet Pytorch
Licence: bsd-3-clause
PyTorch implementation of PathNet: Evolution Channels Gradient Descent in Super Neural Networks
Stars: ✭ 63
Programming Languages
python
139335 projects - #7 most used programming language
Projects that are alternatives of or similar to Pathnet Pytorch
Pathnet
Tensorflow Implementation of PathNet: Evolution Channels Gradient Descent in Super Neural Networks
Stars: ✭ 96 (+52.38%)
Mutual labels: agi, transfer-learning
Spacy Transformers
🛸 Use pretrained transformers like BERT, XLNet and GPT-2 in spaCy
Stars: ✭ 919 (+1358.73%)
Mutual labels: transfer-learning
Qh finsight
国内首个迁移学习赛题 中国平安前海征信“好信杯”迁移学习大数据算法大赛 FInSight团队作品(算法方案排名第三)
Stars: ✭ 55 (-12.7%)
Mutual labels: transfer-learning
Bert language understanding
Pre-training of Deep Bidirectional Transformers for Language Understanding: pre-train TextCNN
Stars: ✭ 933 (+1380.95%)
Mutual labels: transfer-learning
Graphrole
Automatic feature extraction and node role assignment for transfer learning on graphs (ReFeX & RolX)
Stars: ✭ 38 (-39.68%)
Mutual labels: transfer-learning
Bert Keras
Keras implementation of BERT with pre-trained weights
Stars: ✭ 820 (+1201.59%)
Mutual labels: transfer-learning
Weakly Supervised 3d Object Detection
Weakly Supervised 3D Object Detection from Point Clouds (VS3D), ACM MM 2020
Stars: ✭ 61 (-3.17%)
Mutual labels: transfer-learning
Teacher Student Training
This repository stores the files used for my summer internship's work on "teacher-student learning", an experimental method for training deep neural networks using a trained teacher model.
Stars: ✭ 34 (-46.03%)
Mutual labels: transfer-learning
Average Word2vec
🔤 Calculate average word embeddings (word2vec) from documents for transfer learning
Stars: ✭ 52 (-17.46%)
Mutual labels: transfer-learning
Densedepth
High Quality Monocular Depth Estimation via Transfer Learning
Stars: ✭ 963 (+1428.57%)
Mutual labels: transfer-learning
Transferlearning
Transfer learning / domain adaptation / domain generalization / multi-task learning etc. Papers, codes, datasets, applications, tutorials.-迁移学习
Stars: ✭ 8,481 (+13361.9%)
Mutual labels: transfer-learning
Awesome Knowledge Distillation
Awesome Knowledge-Distillation. 分类整理的知识蒸馏paper(2014-2021)。
Stars: ✭ 1,031 (+1536.51%)
Mutual labels: transfer-learning
Deepfakes video classification
Deepfakes Video classification via CNN, LSTM, C3D and triplets
Stars: ✭ 24 (-61.9%)
Mutual labels: transfer-learning
Big transfer
Official repository for the "Big Transfer (BiT): General Visual Representation Learning" paper.
Stars: ✭ 1,096 (+1639.68%)
Mutual labels: transfer-learning
Skin Cancer Image Classification
Skin cancer classification using Inceptionv3
Stars: ✭ 16 (-74.6%)
Mutual labels: transfer-learning
Seismic Transfer Learning
Deep-learning seismic facies on state-of-the-art CNN architectures
Stars: ✭ 32 (-49.21%)
Mutual labels: transfer-learning
Drivebot
tensorflow deep RL for driving a rover around
Stars: ✭ 62 (-1.59%)
Mutual labels: transfer-learning
PyTorch implementation of PathNet: Evolution Channels Gradient Descent in Super Neural Networks. "It is a neural network algorithm that uses agents embedded in the neural network whose task is to discover which parts of the network to re-use for new tasks".
Currently implemented binary MNIST task and CIFAR & cropped SVHN classification task.
Requirements
- Python 2.7
- numpy
- matplotlib
- networkx
- python-mnist
- pytorch
Usage
Install prerequisites:
$ apt-get install python-numpy python-matplotlib
$ pip install python-mnist networkx
And install pytorch: See http://pytorch.org/.
Run with command:
$ python main.py
If you want to repeat experiment:
$ ./repeat_experiment.sh
To check the result:
$ python plotter.py
Modifications
- Learning rate is changed from 0.0001(paper) to 0.01.
Result
Transfer learning of CIFAR10 -> cropped SVHN recorded higher accuracy than cropped SVHN classification accuracy solely (41.5% -> 51.8%, Second figure).
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].