-
https://github.com/lovemefan/telespeech-asr-python
感谢官方团队开源,如果可以的话,希望能够后续能够pr合并
-
https://stackoverflow.com/questions/78859329/fairseq-installation-fails-missing-version-txt-error-during-build-pyhton-3-11-h
```
error: subprocess-exited-with-error
× Getting requirements to bu…
-
```
ValueError: invalid literal for int() with base 10: '0a0'
```
Possibly related to https://github.com/facebookresearch/fairseq/issues/4532
```
This behavior is deprecated in Hydra 1.1 and will…
-
Hi, is there any instruction on multiple nodes multiple GPUs distributed training with hydra train?
https://fairseq.readthedocs.io/en/latest/getting_started.html#distributed-training
The fairseq doc…
-
as of a couple days ago I'm unable to to run colab medley vox. error below:
INFO: pip is looking at multiple versions of hydra-core to determine which version is compatible with other requirements…
-
### 🐛 Describe the bug
I am using pytorch to do multi-GPU distributed training. When I use 1-node 8-GPUs, the training works well. However, when I used 4-nodes 32-GPUs, after N epochs (N can be 1 to …
-
-
您好!
首先,非常感谢您开源了30万小时的预训练模型。我最近正在使用您的模型对粤语和四川话进行微调,过程中遇到了一些问题,想请教您:
1. **预训练语料中粤语和四川话的占比**:请问在您的预训练语料中,粤语和四川话分别占据了多大的比例?了解这一信息对我微调模型的效果评估非常重要。
2. **动态帧和GPU显存利用**:在Wenet中,通过动态帧来指定batch size时,不同…
-
./decode.sh
INFO:__main__:import user_dir: /Users/gao/TeleSpeech-ASR/data2vec_dialect
Traceback (most recent call last):
File "/Users/gao/TeleSpeech-ASR/data2vec_dialect/run_scripts/../infer.py",…
-
If import this lib first like
```py
from torchfcpe import spawn_bundled_infer_model
```
then use
```py
from fairseq import checkpoint_utils # fairseq version 0.12.3, torchaudio version 2.1.2
``…