sanwei111
sanwei111
hello，i met the same issue，have you fix it？could you share me with your mothods
> > hell，in the file of transformer-multibranch-v2，the class of TransformerEncoderLayer--the code are as follow： > > if args.encoder_branch_type is None:#default=None？？？？ > > self.self_attn = MultiheadAttention( > > self.embed_dim, args.encoder_attention_heads, >...
> > thx，what'S the meaning of [attn:1:32:4, dynamic:default:32:4]？could you show some details about the list > > As I mentioned in my last reply, `args.encoder_branch_type` should not be a boolean...
of course ---Original--- From: ***@***.***> Date: Tue, May 24, 2022 09:54 AM To: ***@***.***>; Cc: ***@***.******@***.***>; Subject: Re: [mit-han-lab/lite-transformer] about the global and localfeatures in fig 3 (#35) Same question,...
> great work! > the input is (batch_size, seq_len, channels), is it right that seq_len is a fixed length? why? I found the same question with you, I wonder that...
hello，i have same question with you，how can you fix it finally
> 数据集就qa就行了？要不要给个prompt的形式，比如三元组那种：prompt，input，output
> 一模一样的情况，请问解决了吗 你好，请问怎么解决
excuse me，can you please explain the Class “class RelPartialLearnableMultiHeadAttn(RelMultiHeadAttn):”.I think the cLASS CAN'T MATCH THE FORMULAS IN THE PAPER SO MUCH!