Haichao Yu

One question about Line 314 in meta_neural_network_architectures.py

1

Hi Antreas, Thanks for your excellent work and code. Here is a question about Line 314 in meta_neural_network_architectures.py. Why is `self.weight` always used even if `params is not None`?

Questions about UNet architecture and Dice loss

3

Hello, I have two questions about the UNet architecture and Dice loss function: 1. In the [UNet paper](https://arxiv.org/abs/1505.04597), there is no Relu layer after the final conv. In experiments of...

Is the training stable enough to always converge to the numbers in README.md

2

Hi, Thanks for your efforts. I used your train.sh and only changed stage to 0. The eval sad turns out to be really high (almost never goes below 80) and...

A question on data augmentation part

5

np.random.randint() returns coordinates of (h, w) instead of (w, h). However, in core/data.py, line 42 - 44, the coordinates are regarded as (w, h)? Is it a bug or ......

Q4Q4 post-train Resnet-50 cannot get accuracy 73.3.

2

Hello, Thanks for the great codebase. I tried `Experiment W4A4 + ACIQ + Bit Alloc(A) + Bit Alloc(W) + Bias correction` by `python inference/inference_sim.py -a resnet50 -b 512 -pcq_w -pcq_a...

Some questions regarding the code

5

Hi Shangyu, Great work! I have some questions regarding the code as below - new_meta_hidden_state_dict is empty [here](https://github.com/csyhhu/MetaQuant/blob/master/meta_utils/helpers.py#L13)? - meta_opt_flag is unused in [here](https://github.com/csyhhu/MetaQuant/blob/master/meta-quantize.py#L162)? - what is the setup for...