16 questions
2
votes
2
answers
49
views
Meta-feature analysis: split data for computation on available memory
I am working with the meta-feature extractor package: pymfe for complexity analysis.
On a small dataset, this is not a problem, for example.
pip install -U pymfe
from sklearn.datasets import ...
- 4,111
0
votes
1
answer
199
views
Understanding the results of Transformers Learn In Context with Gradient Descent
I'm trying to implement this paper:
https://arxiv.org/pdf/2212.07677
(Here's their code):
https://github.com/google-research/self-organising-systems/tree/master/transformers_learn_icl_by_gd
I'm ...
1
vote
0
answers
56
views
When training with MAML and Siamese networks, I've encountered issues where the weights aren't updating or the accuracy remains unchanged
I want to train a model for subjective question scoring using ALBERT and a Siamese network, which consists of a bidirectional LSTM and a fully connected layer. During training, I've noticed that the ...
- 17
0
votes
0
answers
37
views
GradientTape.gradient() returns `None` type
following is the code I'm trying to implement for meta learning using maml algorithm on a specific dataset. Inner loop works well, don't know why in the outerloop gradients are None.
def maml(...
2
votes
1
answer
732
views
How to implement Meta learning on the base model(neural network)
I have a complex data in which there are only two features and around 18211 labels(Multi-label regression problem). The two features are categorical data, feature1 has 6 categories(A,B,C,D,E,F) and ...
0
votes
1
answer
207
views
Is this the correct implementation of a MAML model?
I have used CLIP embeddings of image and text as the input and the output is a label ranging from 0 to 5 (6 way label). I tried to make an implemention of this multimodal 6 way classification using ...
0
votes
1
answer
172
views
Too much fluctuation in F1 Score curve during meta training with MAML
I am training VGG11 on a custom image dataset for 3-way 5-shot image classification using MAML from learn2learn. I am encapsulating the whole VGG11 model with MAML, i.e., not just the classification ...
- 724
0
votes
0
answers
130
views
How to combine multiple dataset efficiently to solve using meta learning?
I am solving a meta-learning problem using Reptile Algorithm as used here. I have two datasets. One contains the following classes: iris, pupil, and sclera along with their annotations. Another ...
- 11
0
votes
0
answers
126
views
Vanishing parameters in MAML JAX (Meta Learning)
I am working on an implementation of MAML (see https://arxiv.org/pdf/1703.03400.pdf) in Jax.
When training on a distribution of simple linear regression tasks it seems to perform fine (takes a while ...
0
votes
1
answer
361
views
How to split classes in few-shot classification using CIFAR-10?
I want to train a model that perform a few-shot image classification using CIFAR-10. So I have to train the model with a small amount of classes and use the rest of the classes for the testing. I'm ...
- 11
0
votes
1
answer
859
views
Why is RandomCrop with size 84 and padding 8 returning an image size of 84 and not 100 in pytorch?
I was using the mini-imagenet data set and noticed this line of code:
elif data_augmentation == 'lee2019:
normalize = Normalize(
mean=[120.39586422 / 255.0, 115.59361427 / 255....
- 6,354
1
vote
1
answer
173
views
Meta-learning to find optimal model from pre-trained models in Tensorflow
I have many pre-trained models with a different number of layers (Models are not Sequential). Training data had a shape (1, 1, 103) for these models and output was a class label between 0 and 9.
I ...
- 967
0
votes
0
answers
54
views
Failing to compute gradient in PyTorch
I've been reading this research paper- https://arxiv.org/abs/1908.00413, and trying to implement the code from GitHub- https://github.com/hoyeoplee/MeLU, however, I run into a runtime error while ...
0
votes
0
answers
280
views
How does one use the mean and std from training in Batch Norm?
I wanted to use the means, stds from training rather than batch stats since it seems if I use batch statistics my model diverges (as outline here When should one call .eval() and .train() when doing ...
- 6,354
1
vote
1
answer
407
views
When should one call .eval() and .train() when doing MAML with the PyTorch higher library?
I was going through the omniglot maml example and saw that they have net.train() at the top of their testing code. This seems like a mistake since that means the stats from each task at meta-testing ...
- 6,354