sonica1987

demo.py 144 行： if visualize and self.structure.visualizer: vis_image = self.structure.visualizer.demo_visualize(image_path, output) hot = pred[0, 0, :, :].detach().cpu().numpy() * 255 cv2.imwrite(os.path.join(self.args['result_dir'], image_path.split('/')[-1].split('.')[0] + '_Hot.jpg'), hot) cv2.imwrite(os.path.join(self.args['result_dir'], image_path.split('/')[-1].split('.')[0]+'.jpg'), vis_image)

i need BERT support for mlm

问题解决了，是剪枝模型路径没设对

This may be due to the model I downloaded from HF using an old version of llama.cpp conversion, which seems to have been fixed in the new version https://github.com/ggerganov/llama.cpp/issues/4331

> Can you tell me if reconverting solved your problem? Hello, I have re exported the gguf model using llama.cpp and the issue has not been resolved. Here is the...

> try this template > > ``` > [system](system > You are a helpful assistant.) > user > {{prompt}} > assistant > ``` same problem Not working Whenis not used...

> 看起来令牌没有以这种提示格式出现。试试这个。第一行和最后一行是空的。 > > ``` > > user > {{prompt}} > > assistant > ``` I am using the app in testlight for testing, but there is still an issue....

>   ☹️