-
Hi, great work!! I am comparing the GLA and the attention (from pytorch) in diffusion model training. The GLA seems quite slower than the attention, even worse in the first epoch. Would love to know w…
-
the issue arises in that during incremental building we splice empty sequences into nothing, this causes a problem with position recording because our incremental solution reports the insertion [1]-> …
-
Hi, m3yrin,
Thanks for your code~! I changed it in fairseq framework, however I find the speed is slow.
And the main reason is when doing dynamic programing, torch.gather log_probs is time-consu…
-
Hello,
I am trying to train a new PLC model with [train_plc.py](https://github.com/xiph/LPCNet/tree/master/training_tf2/train_plc.py) and and several things have caused me difficulties:
1) Can I …
-
Hi, I'm new to ctranslate2, and I'm confused about the scores returned by generator.generate_batch() function. What's the coresponding meaning of the scores in the huggingface generate() function?
…
-
Put links to raw and quality trimmed RNA-seq and MBD-BS data on wiki
-
there are many is None or is not None in
https://github.com/huggingface/transformers/blob/master/src/transformers/modeling_bert.py, for example:
https://github.com/huggingface/transformers/blob/mas…
-
Hi, I am following the steps in the Readme to run the model.
My goal is to be able to run the model to be able to provide my inputs. I dont want to train the model.
I did the following:
- Down…
-
@guoweilong
Dear Dr. Guo,
Hello! A few days ago, I reported an error about bam2cgmap of cgmaptools (https://github.com/guoweilong/cgmaptools/issues/51). You suggested I can change BS-Seeker2 to …
-
> Created by **@mschwerhoff** on 2016-02-15 20:15
> Last updated on 2016-02-15 22:23
The postcondition of `test02` fails in Silicon and Carbon if the assertion in the body is commented.
```
#!tex…