-
Pages with any form of specialised CSS display fine on the frontend but display without styling (or with incorrect styling) in Edit+ or within the admin Viper UI. The consequence of this is that user…
-
# Make sure that the projection matrix P has been downloaded from the google drive beforehand to avoid double computation.
# But if the projection matrix P which we provided is not needed, then nothi…
-
请问alphaedit 的 stats 可以直接用llama2的嘛？
如果不可以，请问哪里可以下载llama3的呢？
谢谢~
-
Hi, I'm having some issues running the SERAC methed. I took a closer look at the https://github.com/zjunlp/EasyEdit/issues/261, but the issue is still not resolved
I tried `JackFram/llama-160m`, but…
-
我在按照示例运行run_knowedit_llama2.py文件时发现当前文件目录下并无./hugging_cache/llama-2-7b这个路径，想请问您能否提供您利用KnowEdit这个数据集测试所用的llama-2-7b模型文件？
-
-
非常感谢您的工作，对我很有启发，有两个问题想请教！
# 问题1
使用WISE复现结果（`T=600 or 1000`），编辑速度很慢，请问有什么方法可以加快编辑速度？
数据集ZsRE，编辑模型Llama2-7b-hf
### T=600
` 8%|███████████████▌ …
-
代码中device是0，1，2，3等数字参数，使用的时候应该统一规范为”.to(f"cuda:{device}")“形式
直接使用to(device)形式会报下面的错误
 in pannellum? How to get cent…
vlkpa updated
4 weeks ago
-
Hello, thank you for the fantastic work on this project. I encountered a potential issue with the `ROME` edition when `copy=False` and `sequential_edit=False`. Specifically, during the execution of `B…