dragen1860 MAML-Pytorch issues

dragen1860 / MAML-Pytorch

Elegant PyTorch implementation of paper Model-Agnostic Meta-Learning (MAML)

MIT License

2.28k stars 421 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

训练自己的数据集

#76 qiongjiekucha closed 6 months ago
0
准确率不变

#75 shiyao1999 opened 8 months ago
4
What is the backup file for and what is the reference navie5 in navie5?

#74 kevinwanglu opened 11 months ago
0
About accuracy

#73 feisichen opened 1 year ago
6
Using the Learner object for my project, Loss not behaving at its best

#72 Metabloggism opened 1 year ago
0
debug

#71 swaggyP9527 opened 1 year ago
0
请问模型权值文件在哪里进行保存？代码在哪里？

#70 XingshiXu opened 2 years ago
1
Why the code for Learner is so complicated?

#69 fikry102 opened 2 years ago
1
Why is `for epoch in range(args.epoch // 10000):`

#68 Waterkin opened 2 years ago
2
can you offer us your requirements of environment?

#67 bbruceyuan closed 3 months ago
0
About training and testing

#66 zkk-web opened 2 years ago
6
您好，对于代码有两个问题，请教您一下，谢谢

#65 CUITCHENSIYU opened 2 years ago
3
Asking about inner and outer loop

#64 tranquangchung opened 2 years ago
0
create_graph parameter is False hence first-order MAML?

#63 gunnxx opened 3 years ago
1
about dataset spiltting

#62 raymond00000 opened 3 years ago
0
Incorrect losses_q

#61 zilin129 opened 3 years ago
3
Can you please add a 1-d CNN model to the learner.

#60 zilin129 opened 3 years ago
0
Does the hessian really gets computed?

#59 tanghl1994 opened 3 years ago
1
omniglot dataset download error

#58 vegetarianfish closed 3 years ago
1
why use custom grad clip function?

#57 QasimWani opened 3 years ago
0
Overlap between meta-training and meta-testing?

#56 atseng17 closed 3 years ago
1
Something wrong in the config of the last max pooling layer.

#55 MrDavidG opened 4 years ago
1
something about Computational Graph

#54 kongbia opened 4 years ago
0
how to save and load the model

#53 asker-github closed 4 years ago
0
Question about variable, losses_q

#52 sailist closed 4 years ago
1
Why not "loss_q = sum(losses_q) / task_num" ?

#51 geekyutao closed 4 years ago
2
Performance on Omniglot is slightly lower than paper report

#50 YaxinLi0-0 opened 4 years ago
3
What's the difference between task_num and N-way?

#49 luciaL opened 4 years ago
0
for step, (x_spt, y_spt, x_qry, y_qry) in enumerate(db):

#48 Peterdingpeng opened 4 years ago
1
Some ideas about parameters updating.

#47 zhaoyu-li opened 4 years ago
2
Only use Conv and Linear

#46 CongWeilin opened 4 years ago
1
RuntimeError: Expected object of scalar type Long but got scalar type Int for argument #2 'target'

#45 Peterdingpeng opened 4 years ago
1
why running_mean/running_var: requires_grad must be False?

#44 arxrean opened 4 years ago
0
Some questions about the codes

#43 HongduanTian opened 4 years ago
0
Measure the model performance

#42 kk2491 opened 4 years ago
17
accuracy format

#41 eghouti opened 4 years ago
1
questions with respect to the no_grad()

#40 d12306 opened 4 years ago
1
Conv-ReLU-BN issue

#39 mileyan opened 4 years ago
1
Mutiple GPU

#38 dipanjan06 opened 4 years ago
0
OmniglotNShot

#37 li829 opened 4 years ago
1
Re-sampling tasks after each epoch increases the performance

#36 ShawnLixx opened 4 years ago
3
Result is not deterministic because of iterating over Python dictionary

#35 ShawnLixx opened 4 years ago
3
How to use the trained MAML model for predicting one image at a time?

#34 vatsalsaglani opened 4 years ago
6
run code have a mistakes on db = DataLoader(mini, args.task_num, shuffle=True, num_workers=1, pin_memory=True)

#33 hongshengxin opened 4 years ago
5
2nd Order or 1st Order Approximation?

#32 Vampire-Vx opened 4 years ago
2
learning_rate

#31 dawn2034 opened 5 years ago
0
Should I fine-tune the model on each new task

#30 AaboutL closed 5 years ago
2
Maybe there is some error in learner.py?

#29 crashmoon closed 5 years ago
1
Confused

#28 kimnoic closed 5 years ago
0
why not set bn_training=False when test net on x_qry in finetunning phase

#27 boy-be-ambitious opened 5 years ago
2