tensorboy
tensorboy
@hangzhang-nlp the weights cannot be downloaed.
It has some errors now: ``` python -m llava.serve.test_message_new --model-name LLaVA-7B-v0 --controller http://localhost:10000 Models: ['LLaVA-7B-v0'] worker_addr: http://localhost:40000 A chat between a curious human and an artificial intelligence assistant. The assistant...
> 对于Tag2Text，可以改成batch批量inference的代码，可以极大的增大inference效率。如果只要tagging的输出，可以在generate函数中只输出tagging就停止。 此外，我们新发布的the Recognize Anything Model (RAM)可以任意自定义识别的类别，可以进一步提升tagging inference效率，此功能会在未来发布。 已经发布了吗还是明天发布
Same issues
Hi @bleakie 可以更具体一点怎么改的更好吗，这些参数还没开始深究。。。
谢谢分享，不知道有地方可以参考下您的么。
好的谢谢分享，我试下这个参数。 tensorrt的结果我再看一下，谢谢！
The bug comes from the path of the mask: https://github.com/tensorboy/pytorch_Realtime_Multi-Person_Pose_Estimation/blob/master/training/datasets/coco_data/COCO_data_pipeline.py#L226-L231 And I will fix it now!
The bug is come from the path of the mask: https://github.com/tensorboy/pytorch_Realtime_Multi-Person_Pose_Estimation/blob/master/training/datasets/coco_data/COCO_data_pipeline.py#L226-L231 I will fix it now!
Maybe decrease the batch size will help. :)