search

tianqi-wang1996 / DeepAccident

Code for the benchmark - DeepAccident: A Motion and Accident Prediction Benchmark for V2X Autonomous Driving.
64 stars 9 forks source link

Train the v2x models on DeepAccident with 8 GPUs #27

Open hyh123123 opened 1 month ago

hyh123123 commented 1 month ago

When I use this command: bash tools/dist_train.sh projects/configs/DeepAccidenta_tiny. py 8-- load from/ During the runtime of data/DeepAcccident_data/breakse_tiny -- work-dir work-dirs/DeepAcccident_tiny, an error occurred: Runtime Error: Expected to have finished reduction in the prior iteration before starting a new one This error indicates that your module has parameters that were not used in producing losses. You can enable unused parameter detection by passing the keyword argument 'find_unussed_parameters=True' to 'torch. nn. parallel. DistributedDataParallel', and by making sure all 'forward' function outputs participate in calculating losses. How should this be resolved?

BITwtz commented 1 month ago

Have you ever tried using a single GPU for training, why I can only read 50 frames with a single GPU？ image