Buckets:

Rishik001's picture
download
raw
733 Bytes
"""Run the pipeline on 3000 skatingverse samples and report timing."""
import sys
import time
sys.path.insert(0, "/workspace")
import pipeline
config = dict(pipeline.CONFIG)
config["datasets"] = dict(pipeline.CONFIG["datasets"])
config["output_dir"] = config["output_dir"]
config["skeleton_cache_dir"] = config["output_dir"] / "skeleton_cache"
config["pose_estimator"] = pipeline.POSE_ESTIMATOR
config["yolo_weights"] = pipeline.YOLO_WEIGHTS
config["yolo_imgsz"] = 256
config["max_videos"] = 3000
start = time.perf_counter()
ok = pipeline.run_pipeline(config, sources=["skatingverse"])
elapsed = time.perf_counter() - start
print(f"\n{'=' * 60}")
print(f"DONE — success={ok}, wall time: {elapsed:.1f}s ({elapsed/60:.1f} min)")

Xet Storage Details

Size:
733 Bytes
·
Xet hash:
033f5ba4445837c01cf6f2a1c43354122caa7e52295e963151cddfdc250897da

Xet efficiently stores files, intelligently splitting them into unique chunks and accelerating uploads and downloads. More info.