molokanov50
molokanov50
Unfortunately to share my project is impossible. I'll better repeat in a simpler way. At every `fairseq-interactive` query, where I specify `--input`, `--source-lang` and `--target-lang`, I have to wait some...
I found out everything i need to sample `BatchEncoding` objects as you said - `input_ids` and `attention_mask` attributes need to be sampled correctly. And on the whole, your methodics helped...
@Mycatinjuly No, there is still too little research in this direction.
Update: there is no obvious difference depending on the length of every single text. I reduced input text length down to 10 tokens, or approx. 80 symbols only. Memory usage...
@makaylacmac2 @t1307109256 Put a dot before `/data` since without a dot it is considered as absolute path which obviously doesn't exist in linux. It worked for me.
I trained the model on FF++ c23 and got 82% auc on CDF