QZ
@mchari how did you solve the problem?
@brendan-ai2 I does not work for me, `if os.environ.get("ALLENNLP_DEBUG"): LEVEL = logging.DEBUG else: LEVEL = logging.ERROR`
@yysirs 加载bert模型而不是加载roberta模型就避免了。
> Hi, > > Thanks for your organized repository and for sharing your code. > Your paper about BERT4Rec is interesting. I have a few questions and I appreciate it...
> Hello authors, thank you for your work! > > I am interested in beginning research on document ranking, and I was wondering if it would be possible to get...
> I want too Have you got it ?
@xiotangyuan 好久远的问题，不知道记忆的还对不对。ernie和bert的模型结构是一样的，不同的是预训练任务。
@kiritoyu how to load llama with 13B or 65B, I can load them, but when I give it a prompt, is didn't return any response. 
@yysjasmine Not now. Do you have any idea?
@yysjasmine Sounds great. I am trying the Evaluation framework released with GPT-4, i.e., Evals.