gaosanyuan
gaosanyuan
hi, @cchen156 thanks for your reply. I understood the reason, but it is not the answer I am looking for. I mean if there is some solution to this problem....
ok, I read code in line 213 of common.py file, " num_classes += 1 # always use background as label 0", which may be a misleading
Do you have any conclusions on this question? Thanks I am confused by the same problem. such as as dot line on the road
> ```shell > Building an engine from file gpt2-pretrained.onnx; this may take a while... > [02/19/2022-21:14:43] [TRT] [W] Skipping tactic 0 due to insuficient memory on requested size of 1170309120...
In my opinion, the 'max_classes' and 'max_overlaps' is used to describe the relationship between the proposal roi and the gt roi.
Generally, once a parameter is fixed, it will not be updated in the training process. In mxnet framework, the fixed parameter works as a parameter of the Module class. Please...
@Fenax79 Any ideas for that? I just would like to increase the weights of the larger objects, and I don't know if it should work.
In the current code , the gradient of the mask loss is still: self.factor * mask_weight * (mask_prob - mask_target) / loat(n_rois_fg), and the factor is 1.0/32 In my opinion,...
I found that, the most time is cost for getting the batch data. I tried to prefetch the batch data in multipy process, but it didn't work(still slow). Any other...
> Can you provide some more details? It is impossible to answer your questions without known what machine / environment you are using (e.g., desktop machine, cluster instance, etc.). I...