-
OS: Windows 10 , Anaconda
CUDA_toolkit: 10.1
Python: 3.7
Framework : tensorflow-gpu (2.3.0), pytorch(1.7.0)
Hello. Thank you for your projects
I have a problem with warprnnt
>No module named '…
-
I have been trying to install espnet on my system with default system python, however, I'm getting a CUDA error "fatal error: cuda_runtime_api.h: No such file or directory". I have tried some suggesti…
-
Hi,
The "from warprnnt_tensorflow import rnnt_loss as warp_rnnt_loss" command raise an error:
...warprnnt_tensorflow/__init__.py", line 6, in
_warprnnt = tf.load_op_library(lib_file)
.../load…
-
So, I have tried to implement the model more or less "from scratch" on the basis of this repository.
For that I have implemented a training loop which I am executing eagerly to be able to debug.
…
-
Hello, everyone! I was trying to running train_conformer.py on LibriSpeech dataset(in particular, dev-clean), and i've got this error below:
Run on 1 Physical GPUs
Model: "conformer_encoder"
______…
-
I am following the recipe in this link
https://github.com/espnet/espnet/blob/master/egs2/TEMPLATE/tts1/README.md
Everything runs smoothly but in ```exp/tts_train_raw_phn_tacotron_g2p_en_no_space…
-
Hi, recently I noticed that the loss_rnnt is significantly less than other loss when I tried a multi task Transducer mode. For example:
`iteration` `main/loss_ctc` `main/loss_rnnt` `main/loss_att` …
-
**Conda environments:**
- OS information: 【linux ubuntu 18.04】
- python version: [3.7) [GCC 7.5.0]]
- pytorch version [e.g. pytorch 1.1.0]
**nvcc -V:**
vcc: NVIDIA (R) Cuda compiler dr…
-
When I use espnet/egs/mini_an4/asr1/conf/train_transducer.yaml or espnet/egs/vivos/asr1/conf/tuning/transducer/train_transducer.yaml config to train a RNN-T model, I meet a strange problem:
If ng…
-
Hi everyone! Could you help me with this issue? I tried to apply a lot of answers from other issues but have no found a solution. As I understand it, this problem is related to PyTorch and Cuda versio…