lorinma
lorinma
Hi, do you mind share the code to plot the figure? I found the original notebook but was unable to reproduce it. thanks!
您好!麻烦是否可以给出训练使用了多少卡时的估计吗? 另,法律方向应该由专业人士来做评判,现在有没有效果不好的negative example呢?这并没有在repository里找到。 谢谢!
I use the "lawgpt-lora-7b" as the lora weights and run the webui.sh. When I run the code on cpu, the response is good. When I run the code on gpu,...
您好,关于预训练阶段语料不知道以下几点是否有特殊的考量? 请问法律增量预训练阶段的语料大约是多少B的量级?全都是法律的吗,没有和通用的做混合?以及同属于法律数据,不同数据来源的比例?
Hi, I have created my own dataset but the results looks like this, do you have any idea why so?  The way I prepared my data is to capture...
Hi, many thanks in advance! I'm trying to create a new and complex NL2SQL dataset, however the eval does not seem to support inner join. What should I do except...
你好,json转化成jsonl以后,中文是乱码(类似\u5047\u8bbe\u4f60\u662f),麻烦可以解释一下为什么吗?谢谢
hf上lfs下载了三次都不成功,太不稳定了。谢谢
使用train_multi_gpu, 两张3090显存报OOM。一开始是加载就OOM,把命令行中的FP16去掉后能够训练,但是不久就OOM,显存占用几乎是顶格23.4G/24G。然后我把加载模型的时候去掉了.half()加上了load_in_8bit=True,报错:ValueError: You can't train a model that has been loaded in 8-bit precision on multiple devices. 看了是accelerator不支持的问题。
你好,我无法找到文件: data_path=/wjn/nlp_task_datasets/kg-pre-trained-corpus/total_pretrain_kgicl_gpt,感觉看的有点模糊,麻烦指个路,谢谢!