sandro

Hi, I am in China mainland, and I find it hard for me to download the models via the way in "download_models.sh" file, which might result from the network gap...

Why your PCA components have 512 dimensions rather than 18*512 in Stylegan2 ?

1

I am really confused for this question. Is there anyone can answer it?

Is any plan to open-source your sketch dataset?

2

Currently I cannot find any good enough face sketch dataset on website. It would be helpful for many researchers if you give the source.

How do you train the mfa acoustic model?

1

I follow your tips on training the mfa acoustic model but cannot get labels on aishell3 as accurate as the one you offered. I see there is 'sp' in the...

How many epochs do you use to train your released pretrained model "pretrained_SpanPSP_Databaker.pt"?

如何关闭wandb？我不用nohup的时候可以正常训练，但用nohup跑的时候总是出现以下wandb的error。

2

### Reminder - [X] I have read the README and searched the existing issues. ### Reproduction CUDA_VISIBLE_DEVICES=0 python src/train_bash.py \ --stage sft \ --do_train \ --model_name_or_path /ossfs/workspace/.cache/qwen1.5-0.5b \ --dataset spo_20240430...

### Reminder - [X] I have read the README and searched the existing issues. ### Reproduction 输入： system You are a helpful assistant. user 本产品保险期间 1年 assistant 输出： 1年 Human:...

关于cpp部署

2

你好，想请教一下怎么对原生的python项目做cpp的部署？就像你这样用cpp部署chatglm，我也想对自己的项目进行cpp部署，我只知道一开始需要转换成script model，后面可以用libtorch在c++里面加载，想问一下你是用这种方式的吗？如果不是的话，有什么更好的办法？

请教一下如何在加载模型后实时更新lora？

3

### Reminder - [X] I have read the README and searched the existing issues. ### Reproduction 假设我有1个pretrain的模型、3个针对不同数据集sft的lora参数，请问我可以在加载模型后，根据场景需求，实时替换lora吗？如果可以这样实现的话，可以节省相当一部分机器资源。 ### Expected behavior _No response_ ### System Info _No response_ ### Others _No...