Hrayo

NAS on semantic segmentation

1

Hi, I am wondering if its possible to currently run a NAS algorithm (e.g., SPOS) on a semantic segmentation model from mmsg (e.g., Deeplab). I see that for SPOS, the...

Hi, Thanks for your work. I was wondering how do you deal with gradient updates on the non-searchable stages of the model. The searchable layers will only be updated once,...

Incorrect output shape when using dilations and strides > 1 with depthwise conv

3

### Issue Type Bug ### Source binary ### Tensorflow Version tf 2.4 - v2.4.0-rc4-71-g582c8d236cb 2.4.0 ### Custom Code No ### OS Platform and Distribution Linux Ubuntu 18.04.6 LTS ### Mobile...

Are the weights of shufflenetv2.pth from ImageNet training ?

2

@dog-qiuqiu Are the shufflenetv2.pth weights which are loaded by default in `shufflenetv2.py` from ImageNet training ?

Write Operation Cycles

1

Hello! Hope you're doing well, and sorry to bother you. I am currently using your design for one my implementations. However, I require that the write operation last less than...

Question: Why is Eigendamage and FPGM not docummented nor exposed in importance.py

2

Hi @VainF , first of all, thanks for the great work! Looking at `importance.py` I noticed that there are 2 importances which are not documented in the docs, nor exposed,...

Sudden model size increase for high pruning ratios

1

Hi @VainF First of all, thanks for the amazing work. I am currently working on deriving a tradeoff curve for ViT-b-16 (torchvision) to show the accuracy vs compression tradeoff for...

Hi @VainF, Thanks again for the amazing work! From the Depgraph paper and the source code, I see LSTMs are technically supported. However, I have experienced similar behavior as the...

is LLM-Pruner integrated in Torch-pruning?

1

Hi @VainF , Thanks for the amazing job!. I see some support for pruning LLMs (e.g., QWEN) has been in progress in Torch pruning. But I am also aware of...