-
I have trained transformer using config :
```
conf/tuning/train_pytorch_transformer_d6-2048.yaml
conf/tuning/decode_pytorch_transformer.yaml
```
But got loss-inf warning, and continues like thi…
-
I commented out a test from `test/checker/Makefile.am` that prevents building automatic packages but doesn't reproduce systematically on gentoo or fresh hirsute docker with:
```
FROM ubuntu:hirsute
…
-
## 1. Evaluation of UUV Simulator's Sonar Plugin [Completed]
GazeboROSImageSonar Plugin is well documented with development histories at [image_sonar_description](https://github.com/Field-Robotics-La…
-
**Describe the bug**
I'm running espnet2/commonvoice (german) and the training of the language Model fails with a RuntimeError. In the log file its says:
```
The training was finished at 20 epochs…
-
**Describe the bug**
I'm running espnet2/commonvoice (german) and the training of the language Model fails with a RuntimeError. In the log file its says:
```
The training was finished at 20 epochs…
-
Regarding the implementation of the greedy decoder, I believe the following if statement should be replaced with a while loop:
https://github.com/espnet/espnet/blob/0fe5068a356f5e48beac25117fa1d6c2…
-
I tried to test with "python examples/conformer/train_conformer.py" when I encounted the above error. It happend in line 114:"conformer._build(speech_featurier.shape)" and the shape is [None,80,1], Bu…
-
hello , I found two problems when use TPU
1. /usr/local/lib/python3.6/dist-packages/tensorflow/python/framework/tensor_util.py:445 make_tensor_proto
raise ValueError("None values not s…
-
## Environment
- Development Kit: Internal platform
- Module or chip used: [ESP32-WROOM-32D]
- IDF version:
v3.3-beta1-512-g89ae590
- Build System: [Make]
- Compiler version:…
-
你好：
有一个小小的疑问，
在CPU上训练，linux 16.04，使用aishell_1中的几个人的数据（2100条音频，验证代码用）；训练 ConformerTransducer， 其它参数默认。
```
020-10-05 10:28:11,241 - root - INFO - trainer resume failed020-10-05 10:28:11,241 - ro…