Abstract: It is challenging to annotate large-scale datasets for supervised video shadow detection methods. Using a model trained on labeled images to the video ...
#!/bin/bash set -e # 配置参数 model_path="/d2/models/Qwen3-30B-A3B" train_files="/d2/datasets/verl/gsm8k/train.parquet" val_files="/d2/datasets/verl/gsm8k/test ...
[36m(MegatronPolicyTrainer pid=15469, ip=10.100.0.114) [0m [rank32]:[E830 10:11:36.146373856 ProcessGroupNCCL.cpp:629] [Rank 32] Watchdog caught collective operation ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results
Feedback