Datasets:
Tasks:
Audio Classification
Sub-tasks:
keyword-spotting
Languages:
English
Size:
10K - 100K
ArXiv:
License:
File size: 404 Bytes
fb12e65 eef88e0 fb12e65 72d08a5 fb12e65 c624b60 fb12e65 |
1 2 3 4 5 6 7 8 9 10 11 |
import pandas as pd
_SPLIT = "train"
df = pd.read_parquet("data/dataset_audio_" + _SPLIT +".parquet.gzip")
clipped_df = df.filter(["label_string", "probability", "probability_vector", "prediction",
"prediction_string", "embedding_reduced"], axis=1)
clipped_df = clipped_df.sort_values(by=["label_string"])
clipped_df.to_parquet("data/dataset_audio_" + _SPLIT +"_clipped.parquet.gzip") |