self-gen naming

This commit is contained in:
51616 2025-06-03 15:33:53 +00:00
parent 8b15901fb9
commit 7a889d0d7e

View file

@ -124,7 +124,7 @@ def self_generate(
ds_name = os.path.basename(os.path.dirname(parquet_file))
shard_name = "_" + os.path.basename(parquet_file).replace(".parquet", "")
else:
ds_name = ds_name.replace("self_gen/", "")
ds_name = ds_name.split("/")[-1] # Extract just the dataset name
print(f"Loading dataset: {ds_name} with split: {split}")
kwargs = dict(