Haowen Li

Is GTA-Human II of the download link [https://pjlab-my.sharepoint.cn/personal/openmmlab_pjlab_org_cn/_layouts/15/onedrive.aspx?id=%2Fpersonal%2Fopenmmlab%5Fpjlab%5Forg%5Fcn%2FDocuments%2Fopenmmlab%2Dshare%2Fdatasets%2Fhuman3d%2Fgta%2Dhuman%5Frelease%5F4x&ga=1]?

May I ask how we can save the time of transferring smplx to smpl? The optimization time is too long when the number of data sets is large.

Link to theards.csv https://mega.nz/#F! NtsCGTgD! Links to urXdXLJ6yITYdWEdWN-H1w and comments.csv https://mega.nz/#F! NtsCGTgD! urXdXLJ6yITYdWEdWN-H1w is invalid, could you please provide a new link?

如何使用4bit的模型运行gradio_demo_composition.py

1

使用一张V100运行这个文件会显存不足，因此我想使用4bit模型来进行这个任务。我将`gradio_demo_composition.py`中的170行修改为 `self.model = InternLMXComposer2QForCausalLM.from_quantized(code_path, device_map='auto', trust_remote_code=True).eval()` 但是运行时报错如下： `Traceback (most recent call last): File "/data/intern/lihaowenbj/anaconda3/envs/internlmx/lib/python3.10/site-packages/gradio/queueing.py", line 489, in call_prediction output = await route_utils.call_process_api( File "/data/intern/lihaowenbj/anaconda3/envs/internlmx/lib/python3.10/site-packages/gradio/route_utils.py", line 232, in call_process_api output =...

Can you upload the PyMAF-X_model_checkpoint.pt file?