Janusz Lisiecki

Hi @cnut1648, The video reader operator doesn't support variable frame rate videos. We added a heuristics that should detect this kind of case, still, the heuristic is not 100% accurate...

Hi @linyu0219, DALI can't take the video stream from a webcam as an input, you may want to check [the Deep Stream SDK](https://developer.nvidia.com/deepstream-sdk), however, if you have an image it...

Hi @danbider, DALI 1.14 has been released a week ago so it could be it. Can you tell us how exactly we can reproduce that problem?

I see that also [a new version of pytorch-lightning](https://github.com/PyTorchLightning/pytorch-lightning/releases/tag/1.6.4) has been released, so it could be the second thing that may cause problems.

Hi @ksm9413, Please check [the `as_array` function documentation](https://docs.nvidia.com/deeplearning/dali/user-guide/docs/data_types.html#nvidia.dali.backend.TensorListCPU.as_array). To call it, all tensors in the batch need to have the same shape, otherwise the bach cannot be represented as a...

Hi, @Data-drone, The file reader operator reads single files, while the webdataset is a container. You can read more about it [here](https://webdataset.github.io/webdataset/gettingstarted/). The class is just one of the data...

Hi, @Data-drone, You can use [the `get_property` operator](https://docs.nvidia.com/deeplearning/dali/user-guide/docs/operations/nvidia.dali.fn.get_property.html). As an example you can check [this](https://github.com/NVIDIA/DALI/blob/de3d5a8abb6217c7c7dc1a501e516103f37a8a3a/dali/test/python/operator/test_get_property.py#L30) test file.

Hi @KsawerySmoczynski, I'm not sure if you should recreate the DALI pipeline in every call to the `*_dataloader` method. Can you check [this example](https://docs.nvidia.com/deeplearning/dali/main-user-guide/docs/examples/frameworks/pytorch/pytorch-lightning.html), and create it only once in...

Hi @KsawerySmoczynski, > As you can see there is a almost 10 minutes break between init of 1-st train epoch and batches being processed (line 98 and 99). Then the...

Hi @KsawerySmoczynski, I have tested `PYTHONUNBUFFERED=1` and it helped me in the case I had. Maybe here there is a different cause of the problem. I run your repro (with...