-
Hi, I'm getting this dll import errors when I'm trying to import torch since 1.4 if I use Python 3.8 for build. The error below is from source build in a virtualenv that I tried today. I can build and…
-
**Describe the bug**
Not a bug, ask for help and training techniques
conformer-ctc small can't converge on Librispeech 960h
**Basic environments:**
- OS information: Ubuntu 18.04.1 LTS
- pyt…
-
I was using Marlin 1.0.2 and i got this issue, then I updated to Marlin 1.1.0-RC see i…
-
Hi,
I am trying to run the vivos asr1 recipe in espnet1 with the following train.yaml config file(it's the default file given):
batch-size: 30
maxlen-in: 800
maxlen-out: 150
criterion: loss…
-
## Environment info
- Platform: Ubuntu 16.04
- Python version: 3.7.10
- PyTorch version (GPU?): PyTorch 1.7.0, CUDA 10.2
- Using GPU in script?: GeForce RTX 2080 Ti
## Information
Model …
-
**Hardware**
Teensy 4.1
CC1101
uBlox M8N GPS without configuration flash memory (compass is on the way)
**Problem**
my T110 Display shows only two messages if they are activated on GPS, if all …
-
**Describe the bug**
I am training an Arabic model with [diacritics](https://en.wikipedia.org/wiki/Arabic_diacritics).
Digitally, each diacritic is represented as a separate (unicode) character fr…
-
@b-flo I think I meet a bug in VIVOS.
Pytorch: 1.9
CUDA: 11.1
The script is running with all default setting and I just add two line as followed in to train.yaml.
`
noam-lr: 5.0
noam-adim: 4…
-
### Question
transducer loss is better than ctc in paper https://arxiv.org/pdf/2011.04785.pdf. it's use LC-BLSTM as encoder to compare 3 loss ,have you try to train a TDS-transducer, is it better th…
-
Environment:
- Ubuntu 20.04
- TensorFlow 2.3
- TensorFlowASR main
Command:
```
python examples/conformer/train_subword_conformer.py --tbs 8 --ebs 8 --mxp --devices 0 --cache --subwords ./outpu…