CanyonWind Single-Path-One-Shot-NAS-MXNet issues

CanyonWind / Single-Path-One-Shot-NAS-MXNet

Single Path One-Shot NAS MXNet implementation with full training and searching pipeline. Support both Block and Channel Selection. Searched models better than the original paper are provided.

151 stars 22 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

Supernet training is too slow

#18 Spark001 opened 4 years ago
0
Export ONNX model

#17 soyebn opened 4 years ago
1
Supernet Training with Constraints

#16 betterhalfwzm opened 4 years ago
0
Supernet model weights

#15 kepengxu closed 4 years ago
1
supernet training with resource constrain

#14 pawopawo closed 4 years ago
2
the loss of supernet can't converge

#13 cavalleria closed 4 years ago
5
Add cpu quantized model

#12 ZhennanQin closed 4 years ago
0
Context switching causes multi GPU idling

#11 CanyonWind opened 4 years ago
2
Official Review

#10 micronet-challenge-submissions closed 4 years ago
14
Develop

#9 CanyonWind closed 4 years ago
0
Bug in random_channel_mask

#8 xieydd closed 4 years ago
4
Channel selection is disabled after resuming

#7 CanyonWind closed 4 years ago
1
Can you offer a toy data and example like "CIFAR10"?

#6 xuehui1991 closed 4 years ago
3
evolution search

#5 pawopawo closed 4 years ago
4
supernet training

#4 pawopawo closed 4 years ago
15
the loss of supernet doesn't seem to converge

#3 pawopawo closed 4 years ago
11
Update train_imagenet.py

#2 yaooxii closed 5 years ago
0
Merge fixed structure net and none-fixed random selected structure net

#1 CanyonWind closed 5 years ago
0