aohenuo's github account
aohenuo's github account
you can use this code to convert the txt to pictures into folds ： ''' import matplotlib.pyplot as plt from PIL import Image pic_x = np.loadtxt('G:/the experiment of autumn school/NLP/output0_trn_img3.txt')...
+1
理论上来说，对于现有所有大型预训练语言模型，都可以用修改history或者是之前文本来让其输出某些你想要的内容（毕竟本质就是对于前文对话建模罢了）
> Hey [@aohenuo](https://github.com/aohenuo), > > I'm unable to reproduce this behavior with a smaller example. Could you, similarly to [huggingface/transformers#36509](https://github.com/huggingface/transformers/issues/36509), reduce your snippet a bit to reproduce the error with...
I've implemented my own version of prefix tuning, so I no longer need to use PEFT's prefix tuning. Thank you all very much!