trangtv57
trangtv57
@rikrd can you share your solution for fix cuda OOM?
tks @danpovey , I will check then report result later.
hi @vineelpratap as some comment above, problem just occurred when language model not enough good meaning fit with audio domain. So as this idea, i just use lm_3 gram with...
hi @abhinavkulkarni 1. Layernorm with time or not will be change the result of model (because pretrained model has trained with layernorm with time) but you can use it (layernorm...
hi @abhinavkulkarni, you can easy handle language model build by kenlm with limited vocab like script preprocess in lexiconfree-model. [script to build limited vocab for kenlm](https://github.com/facebookresearch/wav2letter/blob/master/recipes/lexicon_free/librispeech/prepare_fairseq_data.sh) and [script for building...
sry bạn, drive của mình full lưu trữ nên mình lỡ tay xóa mất checkpoint, bạn có thể làm theo hướng dẫn để train lấy check point như mình ghi...
i have same issuse, pls, fast check this error :(
sorry but any update @pkufool pk?
tks @pkufool, i will check it then feedback you soon.
> I can successfully run this repo by using pytorch==1.13.1 and xformers==0.0.16  can you give me the command to install pytorch and xformers right version, i try install with...