Audio transcription

In [1]:

from indexify import IndexifyClient
client = IndexifyClient(service_url="http://100.114.163.163:8900")

from indexify import IndexifyClient client = IndexifyClient(service_url="http://100.114.163.163:8900")

!pip install pandas

In [78]:

import pandas
data = pandas.read_csv("~/Downloads/oral-arguments-2022-08-03.csv")
len(data)

import pandas data = pandas.read_csv("~/Downloads/oral-arguments-2022-08-03.csv") len(data)

/var/folders/lr/j5bzp2894pq9hv33bhl4hkhr0000gn/T/ipykernel_13820/3150347646.py:2: DtypeWarning: Columns (6) have mixed types. Specify dtype option on import or set low_memory=False.
  data = pandas.read_csv("~/Downloads/oral-arguments-2022-08-03.csv")

Out[78]:

79719

Install indexify -

curl https://www.tensorlake.ai | sh

indexify server -d

Download Whisper extractor -

indexify-extractor download hub://audio/whisper-asr
cd indexify-extractor
indexify-extractor join-server whisper:WhisperExtractor

Or run with docker

docker run -d -v /tmp/indexify-blob-storage:/tmp/indexify-blob-storage -p 9500:9500 tensorlake/whisper-asr join-server --workers=1 --coordinator-addr=host.docker.internal:8950 --ingestion-addr=host.docker.internal:8900 --advertise-addr=0.0.0.0:9500

In [81]:

data = data[["download_url", "docket_id"]]
non_empty_download_urls =data[data["download_url"].str.startswith("http")]
print(len(non_empty_download_urls))
unique_urls =  non_empty_download_urls.drop_duplicates(subset=['download_url'])
len(unique_urls)

data = data[["download_url", "docket_id"]] non_empty_download_urls =data[data["download_url"].str.startswith("http")] print(len(non_empty_download_urls)) unique_urls = non_empty_download_urls.drop_duplicates(subset=['download_url']) len(unique_urls)

79719

Out[81]:

79539

In [127]:

for index, row in unique_urls.head(100).iterrows():
    client.ingest_remote_file(url=row["download_url"], mime_type="audio/mpeg", labels={"docket_id": str(row["docket_id"])})

for index, row in unique_urls.head(100).iterrows(): client.ingest_remote_file(url=row["download_url"], mime_type="audio/mpeg", labels={"docket_id": str(row["docket_id"])})

In [110]:

client.add_extraction_policy(extractor='tensorlake/whisper-asr', name="audio-transcription")

client.add_extraction_policy(extractor='tensorlake/whisper-asr', name="audio-transcription")

In [ ]: