-
I tried to make some modifications to diff-gaussian-rasterization, adding a backpropagation. However, an error will be reported halfway through the operation. I would like to know how to open the gene…
-
I didn't understand this loss functoin `return -dice_coef(y_true, y_pred)`. For backpropagation I think we need a differentiable loss function for instance `return 0.5*math.pow(1-dice_coef(y_true, y_p…
-
Hello, I am a graduate student. I tried to execute my project based on torch_ACA solver in multiple gpus with DP wrapper recently. However, I found that the performance will decrease much compared wit…
-
The source code of DeepSpeedEngineWrapper:
```
class DeepSpeedEngineWrapper:
"""
Internal wrapper for deepspeed.runtime.engine.DeepSpeedEngine. This is used to follow conventional training…
-
Trying to read your example code without comments or the documentation handy in a browser window is a pretty puzzling experience. Lots of numeric values without a hint as to what that value might be. …
-
Hello Author @skokec ,
I am reading your paper and trying to understand the relevant codes, such as the translation of function in the paper, but the large number of cuda operations and some related…
-
# Improvements for overall network performance
- [x] Backpropagation in Batch (currently use batch size = 1)
- [ ] Batch normalization:
> $$N(x) = \frac{(x - \mu)}{\sigma} \cdot \alpha + \beta$$
…
-
I thought add these features are needed for all types of neuron networks, so it would be nice to add them.
In addition, maybe we could switch between feedforward and backpropagation mode?
-
I want to optimize the placement of an object. But when I try this , I failed to get the gradient of the object's position.
Then I find this is because the object's position is not trainable.
I want…
-
Hi,
I am currently trying to recreate the results in the paper. However, I am unclear on why the mask array is flipped along the batch and filter dimension using END : -1 : 1 in backpropagation. I …