returnn
returnn copied to clipboard
rwth-i6
The RWTH extensible training framework for universal recurrent neural networks
This PR fixes #1087. As I face the issue in the context of supervised multilingual training, I added a more general test case also for that which does not necessarily...
Hi! I'm trying out using returnn_common together with the hybrid pipeline from https://github.com/rwth-i6/i6_experiments/blob/main/common/setups/rasr/hybrid_system.py. I think I have config generation using the serializer from i6_experiments working. However, for decoding RASR needs...
I have a case where I'd like to use the GatherLayer on the batch axis. This works, but the size placeholder is not gathered. Does anything speak against gathering on...
I have a corpus with segments that have `````` attached to them. When I want to use it in an ExternSprintDataset (e.g. with dump dataset), the process crashes with ```...
I am right now reproducing a config which is using ```python 'phoneme_probs': {'class': 'softmax', 'from': ['spectrogram_encoder'], 'target':['data:classes'], 'loss': 'ctc'} ``` as a loss layer. Since I am using returnn_common I...
Currently it is mixed, via `Dataset.iterate_seqs` (chunking logic) and `Dataset._generate_batches` (batching logic), although the logic is already mostly separate, and in principle, it should be separate. (Historically, it might have...
In the [Volta Tensor Core GPU Achieves New AI Performance Milestones (2018)](https://developer.nvidia.com/blog/tensor-core-ai-performance-milestones/) blog post by Nvidia, it is said that NHWC performs best. It sounds like this statement is specifically...
When using the config in: https://gist.github.com/robin-p-schmitt/41da5e1274ccb93be22881f2f1fe91ba, I have the problem that the model does not converge at all. When looking at the learning rate file: ``` 1: EpochData(learningRate=0.0001, error={ 'dev_error_ctc':...
This is just to start some discussion, specifically what we want to drop. Surely = (3,6)` somewhere. * I can't imagine such a situation now, but maybe there could be...