Training

Pre-train

CUDA_VISIBLE_DEVICES=<GPU> python main.py --config cfgs/spike_pretrain.yaml --exp_name --exp_name <name>

Classification on ModelNet40

CUDA_VISIBLE_DEVICES=<GPU> --scratch_model --config cfgs/spike_finetune_modelnet.yaml --exp_name <name>

Classification on ScanObjectNN

# 0. ScanObjectNN OBJ-BG
CUDA_VISIBLE_DEVICES=<GPU> python main.py --scratch_model --config cfgs/spike_finetune_scan_ogjbg.yaml --exp_name <name>

# 1. ScanObjectNN OBJ-ONLY
CUDA_VISIBLE_DEVICES=<GPU> python main.py --scratch_model --config cfgs/spike_finetune_scan_objonly.yaml --exp_name <name>

# 2. ScanObjectNN PB-T50-RS
CUDA_VISIBLE_DEVICES=<GPU> python main.py --scratch_model --config cfgs/spike_finetune_scan_hardest.yaml --exp_name <name>