Yunze Man comments

Results 7 comments of


                                            Yunze Man

Change object storage folder

Try using a soft link to point .objaverse/ to your target location. This is my workaround.

Testing Using Pretrained Model Results in Different Performance Across Multiple Runs

These are test runs so .log was not generated. But the outputs are here: ~/conda-envs/tr3d_env/lib/python3.8/site-packages/MinkowskiEngine/__init__.py:36: UserWarning: The environment variable `OMP_NUM_THREADS` not set. MinkowskiEngine will automatically set `OMP_NUM_THREADS=16`. If you want...

Testing Using Pretrained Model Results in Different Performance Across Multiple Runs

Regarding randomness, I strictly followed your steps. I checked torch randomness guide but didn't find useful directions. Setting `--deterministic=True` didn't seem to help. I also trained the model following your...

Testing Using Pretrained Model Results in Different Performance Across Multiple Runs

Thanks for pointing that out. However, after changing `quantization_mode` of x to `UNWEIGHTED_AVERAGE`, the little randomness still persists. Here are results of three separate runs: Run 1: ([email protected]: 0.7068, [email protected]:...

Content in outputs gradually disappear when Training depth conditioined model

@Tangshitao I observe exactly the same problem. Very curious about the reason.

无法训练

> Sorry to disturb, I got an error when train this network. > Exception: Layer 'IRCNN/in' already exists, please choice other 'name' or reuse this layer. > Could you please...

Release of Training and Evaluatiing process

Will find time to complete the codebase after the cvpr deadline