Caelyn
Caelyn
When I used the command line of cd-hit, I met the same issue. Did you fix the problem?
I also met this issue, and I didn't find the problem sequence like you did. Do you know how to fix it? Thanks a lot.
How do you find the problem sequence?
> > How do you find the problem sequence? > > Actually it's not an efficient way. I separate the original file into small parts with Seqkit software, and ran...
> > > > How do you find the problem sequence? > > > > > > > > > Actually it's not an efficient way. I separate the original...
Hi, I found the problem sequences using the method of you. It seems the problem sequence is too long that the kofamscan cannot handle on it. But I don't know...
> Hey folks, I hit this too. > > (First, thank you to the developers and maintainers of kofamscan :)) > > In wondering why this came up only recently...
Me too! Is there any official documentation about the definition of e-value and score in the output "detail" file?
Hi, thanks for the wonderful work. May I know when the dataset will be released? I only found one example dataset. Thank you in advance.
I also noticed the change. The reason for the change can be found in: https://github.com/volcengine/verl/pull/3287. I think it was fixed for multiple-turn dataset. And I agree "loss_mask = batch.pop("loss_mask")[:, 1:].reshape(-1).to(self.device_name)"...