Riga27527

> You may refer to the supplementary to find the links to all meshes. However, I have made several modifications to the original mesh. I'll upload the blender scripts for...

Hi @dorverbin , sorry to bother you. I want to confirm a simple question, the directional MLP width in your article is 256, right? But I found that the default...

> I bypassed the OOM by simply transfering the chunks in `render_image` to CPU. Hi, @GCChen97 , I have encountered the same problem as you. Could you share your code...

> > Hi, @Riga27527 , here is my solution: > > ``` > > def move_chunks_to_cpu(chunks): > > chunks_new = [] > > device_cpu = jax.devices("cpu")[0] > > for chunk...

Hi, @jonbarron . I'm now using blender_refnerf.gin, but due to memory limitations, the batch_size is set to 2048, how many rounds of training would be better? Is 500,000 rounds enough?