add run command

eval.py

@@ -0,0 +1,85 @@
+import argparse
+from multiprocessing import cpu_count
+import os
+from pathlib import Path
+
+from data_loader.videotext_dataloader import create_datasets, create_loaders
+from trainer.trainer import TrainerVideoText
+from utils_collection.general import load_config, set_seed, print_csv_results
+from utils_collection.general import create_dataloader_path
+
+
+def main():
+ parser = argparse.ArgumentParser(description=__doc__)
+ parser.add_argument('config', type=str, help='Experiment to run')
+ parser.add_argument('--checkpoint', type=str, help='Checkpoint to load')
+ parser.add_argument('--group_k',
+ type=int,
+ default=5,
+ help='number of segments per video')
+ parser.add_argument("--modality",
+ type=str,
+ default="howto100m",
+ help="select the modality name")
+ parser.add_argument('--workers',
+ type=int,
+ default=None,
+ help='set number of workers (default #CPUs - 1)')
+ parser.add_argument('--log_dir',
+ type=str,
+ default="runs/eval",
+ help='directory to save/load the runs and logs')
+ parser.add_argument("--val_split",
+ type=str,
+ default="test1k",
+ help="Select the validation split")
+ parser.add_argument("--dataroot",
+ type=str,
+ default="data",
+ help="change datasets root path")
+ parser.add_argument("--cuda", action="store_true", help="train on GPUs")
+ parser.add_argument("--single_gpu",
+ action="store_true",
+ help="Disable multi-GPU")
+ parser.add_argument("--preload_vid",
+ action="store_true",
+ help="Load video features into RAM")
+ args = parser.parse_args()
+
+
+ configuration = load_config(args.config)
+ set_seed(0)
+ num_workers = min(10,
+ cpu_count() -
+ 1) if args.workers is None else args.workers
+
+ # meta_data_path = Path(os.path.join(args.dataroot, "meta", "meta_group{}.json".format(args.group_k)))
+ # video_feat_path = Path(os.path.join(args.dataroot, "group{}".format(args.group_k), "video_features", "howto_h100m.h5"))
+ # language_feat_path = Path(os.path.join(args.dataroot, "group{}".format(args.group_k), "language_features", "text_default.h5")) 
+ # meta_text_len_path = Path(os.path.join(args.dataroot, "group{}".format(args.group_k), "language_features", "text_lens_default.json"))
+
+ # data_path_dic = {
+ # "meta_data": meta_data_path,
+ # "video_feats": video_feat_path,
+ # "language_feats": language_feat_path,
+ # "meta_text_len": meta_text_len_path,
+ # "dataset_name": configuration.dataset.name
+ # }
+
+ data_path_dict = create_dataloader_path(args.dataroot, args.group_k, configuration.dataset.name, video_feature_name=args.modality)
+
+ configuration.dataset.val_split = args.val_split
+ val_set = create_datasets(data_path_dict, configuration, args.preload_vid, True, eval=True)
+ val_loader = create_loaders([], val_set,
+ configuration.training.batch_size,
+ num_workers, eval=True)
+
+ trainer = TrainerVideoText(args.log_dir, configuration, args.cuda, args.cuda
+ and not args.single_gpu, args.checkpoint, False)
+ trainer.validate(val_loader)
+ trainer.close()
+
+
+
+if __name__ == '__main__':
+ main()

run_scripts/run_crossclr.sh

@@ -1,4 +1,4 @@
-python3 train.py --config_file config/lsmdc_100m_base.yaml --cuda \
+python3 train.py --config_file config/lsmdc.yaml --cuda \
 --dataroot /mnt/efs/fs1/workspace/experiments/data/lsmdc16/modality_experts/ \
 --data_pickle /mnt/efs/fs1/workspace/experiments/data/lsmdc16/modality_experts/modalities_pickle_v2/ \
 --group_k 5 --exp_group mod_mod_combine \

run_scripts/run_toy.sh

@@ -0,0 +1,2 @@
+#python3 train.py config/yc2_100m_coot.yaml --cuda --log_dir runs/lsmdc_toy --dataroot /mnt/efs/fs1/workspace/experiments/h100m_lsmdc/data/howto100m_lsmdc/ --preload_vid
+python3 train.py --config_file config/lsmdc_100m_base.yaml --cuda --dataroot data/lsmdc/toy/ --group_k 5 --exp_group patching --exp_name debug6 --num_runs 2