Yesterday
-
Update README: schema, splits, dedup note, label table
-
Remove duplicate _x videos (1069 rows)
Dropped 1069 train rows whose filename stem matches _\d+$ (bit-identical copies of base videos); updated README train count 3875 -> 2806. -
Upload README.md
-
Upload label_mapping.json
-
Commit shard shard_5-private_test.parquet
-
Commit shard shard_4-private_test.parquet
-
Commit shard shard_3-private_test.parquet
-
Commit shard shard_2-private_test.parquet
-
Commit shard shard_1-private_test.parquet
-
Commit shard shard_0-private_test.parquet
-
Commit shard shard_3-public_test.parquet
-
Commit shard shard_2-public_test.parquet
-
Commit shard shard_1-public_test.parquet
-
Commit shard shard_0-public_test.parquet
-
Commit shard shard_7-train.parquet
-
Commit shard shard_6-train.parquet
-
Commit shard shard_5-train.parquet
-
Commit shard shard_4-train.parquet
-
Commit shard shard_3-train.parquet
-
Commit shard shard_2-train.parquet
-
Commit shard shard_1-train.parquet
-
Commit shard shard_0-train.parquet
-
initial commit
