penglo

Great work, looking forward to your open-source release!

Congratulations on your successful work! Your achievements have been very inspiring to me. I've encountered some issues with visualization due to modifications I made to the model, and I'm wondering...

代码问题

1

Traceback (most recent call last): File "train.py", line 182, in main() File "train.py", line 102, in main loss_out = loss_fn(out, data) File "/root/miniconda/envs/simpl/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1130, in _call_impl return forward_call(*input, **kwargs)...

关于并行计算

1

你好，恭喜您取得如此优秀的成果，我在复现你的代码的时候也是使用了8块3090，但是还是跑不起来您所使用的batchsize ，我研究了一下您使用的多卡训练的代码采用的是数据并行的方式，但是大多数我们用来解决多卡运行显存不足时，通常使用的是模型并行的方式，想请教一下您是否做过这方面的部署呢，如果有，希望您开源一下代码，或者我们进行一些交流，我的邮箱是[email protected]。 Hello,Congratulations on achieving such excellent results. While reproducing your code, I also used 8 RTX 3090 GPUs, but I still couldn't manage to run the batch size you...