-
general domain 코퍼스로 학습된 RoBERTa-large 모델은 이번 프로젝트에서 학습에 사용하는 국회 회의록과 다른 특성(domain)을 가진다고 생각해 unlabeled 회의록 코퍼스로 추가 학습하고자 한다.
특정 domain, task의 데이터로 추가 학습하는 방법의 효과에 대한 논문(**Don't Stop Pretraining**)은 …
-
Hi,
Thanks for the great work!
The code for imagenet finetuned does not seem to work as it can not load a model from VLCTransformer.
I think that is the initial version for MAE.
I wonder …
-
Dạ em chào anh ạ. Em có đang làm một cái project có sử dụng vietocr và gặp một số lỗi ngay trong bước `train`
 worked. I upgraded astropy, and now it fails. Although I'm not exactly sure why. Error message below. Nothin…
-
Hi all, my name is João França. I am a PhD student from CBPF, Brazil and I plan to write my thesis about strong lensing systems. PyAutoLens have been helping a lot! By the way, I am trying to fit 20 d…
-
### Description
Testing Debian build for 5.1RC1
When running tests (during build time) on i386 (32 bit), the tests test_models_fitting[LevMarLSQFitter-model5] and test_models_fitting[LevMarLSQF…
-
Hello,
First of all, well done and thank you for this great work.
I am trying to launch iBOT experiments but struggle with loss going to NaN after a few epochs. The training loss increases (see…
-
We are training Vav2Vec using the run_speech_recognition_ctc_bnb.py-script.
This is working fine with Common Voice, however using our custom dataset and data loader at [NbAiLab/NPSC]( https://huggi…