-
Hi all,
I am trying to use DEQ models to fit 1 toy data point by wrapping the neural net in a SteadyStateProblem. However, when I run a piece of code like this
```
using Flux
using DiffEqSensi…
-
Hi all,
I have just started to learn about this fabulous project. But I have experienced a brick wall with respect to the examples. I try running train_and_evaluate notebook from the examples folde…
lefig updated
4 years ago
-
Hi,
I was going through imaml.py example code that you have given.
I didn't understand the benefit of passing meta_model.parameters() and params as arguments to the inner_loop function [here](http…
-
@mwsht
According to this [article](https://arxiv.org/pdf/1406.2572v1.pdf), the gradient of SCSCNN model will be reeeeeeeeeally flat and full of saddle points...
Maybe consider a alternate trainin…
D0048 updated
4 years ago
-
Should we expect nt.predict.gradient_descent to fail when using jax vmap due to the scipy ode solver? Are there any suggested workarounds for speeding this function up over batches?
-
Armadillo has been manually installed.
[jw@cn06 build]$ [jw@cn06 build]$ make VERBOSE=1 -j1
/usr/bin/cmake -S/data/jw/mlpack -B/data/jw/mlpack/build --check-build-system CMakeFiles/Makefile.cmake…
-
以下のようなコードを実行したところ、`predict`の出力が直感と合わない結果になりました。
```
require 'rumale'
require 'numo/narray'
require 'numo/gnuplot'
require 'numo/linalg/autoloader'
include Numo
x = DFloat[[1, 2, 3, 4, 5, …
-
I am using tensorflow-gpu 1.13.1 and Keras 2.2.4, both of which are up to date. I am trying to make a binary classifier to classify files as benign or malicious, however I keep getting an error.
…
-
Hi!
Somewhat recently I got a new training server which is really fast, but I'm currently having trouble utilizing it's GPU and CPU to it's full potential when training my model.
I'm training an…
-
파일 목차
1. `YOLOv2.py`
2. `darknet.py`
3. `train.py`
4. `loss.py`
5. `config.py` (?)