Peipei Guo

why this pr has so many files changed?

Is 5 seconds too long as the default? Should we reduce `normal_timeout` a bit so that the reconnection and re-push can quickly recover? `TRANS_BY_GPT3`

I am also waiting for this, hope some requests can be save as favorites and use it again when start vscode next time.

> You can fix that by setting the environment variable TCNN_CUDA_ARCHITECTURES=75 and then reinstalling tinycudann. I have encountered a similar problem "OSError: Could not find compatible tinycudann extension for compute...

I feel the quality of result is not good, I also want to know how can we make it better? This is my test result of lego: 

> I am looking into potential bugs in the pipeline that may have also caused several other issues reported. I will update once confirmed. Thanks. Waiting for your good news.