-
Hi,
I'm trying to use ground truth poses from a dataset ([ICL NUIM](http://redwood-data.org/indoor/dataset.html)) directly with a custom dataloader but with no success so far (see screenshot below…
-
First of all congrats on the paper and thank you for making your code public!
I would like to retrain the network on a custom dataset with a custom dataloader. However, I could not find in your code …
-
Hi @weixmath
I am trying to train the model on a custom dataset with 4 classes and 12 views. I am getting the below error:
. I have followed your suggestions for registering the data and using instance_coco_custom_dataset_mapper.py for it. Datal…
-
i run my codes on [RTX2080ti*6] system
1. when i train on my custom dataset with a single gpu, gpu utils(30~40%) are low.(even when i set pin_memory = True) my training command is below. How can i …
-
I’m encountering a memory issue during the **testing** phase of a semantic segmentation task on a custom dataset. The training phase proceeds without any issues, but during testing, the memory usage c…
-
### 🚀 The feature, motivation and pitch
## Pitch
I want to be able to access multiple times my dataset without knowing in advance the indices queried.
I propose to build a `DynamicLoader` inh…
-
Hi, I have created a variation of the PyTorch NLP tutorial https://pytorch.org/tutorials/beginner/text_sentiment_ngrams_tutorial.html so to work with a custom dataset (a `pandas` dataframe), and you f…
-
I wanted to try to train a new model using my own dataset, and when using NLL_ours as the Loss function, the loss value would gradually become negative during training. While training is normal when u…