adrienle / animl Goto Github PK
View Code? Open in Web Editor NEWReproduction of "Model-Agnostic Meta-Learning" (MAML) and "Reptile".
License: MIT License
Reproduction of "Model-Agnostic Meta-Learning" (MAML) and "Reptile".
License: MIT License
Thanks for your ANIML codes!
I tried it using Mac, and found an error when executing fit_transfer(). By debugging, net.train() triggered an error ocurred while starting the kernel, and below is the error message.
OMP: Error #15: Initializing libiomp5.dylib, but found libiomp5.dylib already initialized.
OMP: Hint This means that multiple copies of the OpenMP runtime have been linked into the program. That is dangerous, since it can degrade performance or cause incorrect results. The best thing to do is to ensure that only a single OpenMP runtime is linked into the process, e.g. by avoiding static linking of the OpenMP runtime in any library. As an unsafe, unsupported, undocumented workaround you can set the environment variable KMP_DUPLICATE_LIB_OK=TRUE to allow the program to continue to execute, but that may cause crashes or silently produce incorrect results. For more information, please see http://www.intel.com/software/products/support/.
After googling, I tried:
import os
os.environ['KMP_DUPLICATE_LIB_OK']='True'
Now, it works for me now :)
I notice a function 'V' when defined method copy
of class ModifiableModule
, but cannot find what it mean, nor where it defined. So what it is? I will be grateful for anyone's directions!
โ the V in code
def copy(self, other, same_var=False):
for name, param in other.named_params():
if not same_var:
param = V(param.data.clone(), requires_grad=True)
self.set_param(name, param)
Sorry to bother you because I'm a true beginner in meta-learning. In your ipynb file MAML section, since new model's parameter is same variable to the original model, so there is twice back propagation happening to the original model. I think the second time bp's grad is what we need. The first bp only help to new model. I don't know if i'm right. Or as the perfect original ipynb result shows, your solution is also feasible?
If I was right, I got trouble in revising this mistake for the situation where the batch_size is not 1.
A declarative, efficient, and flexible JavaScript library for building user interfaces.
๐ Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. ๐๐๐
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google โค๏ธ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.