-
Notifications
You must be signed in to change notification settings - Fork 30
Open
Description
I'm having issues with both of them going together
PIIL is good if I set off the speech_understanding
import requests
import time
base_url = "https://api.assemblyai.com"
headers = {
# Replace with your chosen API key, this is the "default" account api key
"authorization": "<>"
}
# URL of the file to transcribe
FILE_URL = "https://cdn.assemblyai.com/upload/<>"
# You can set additional parameters for the transcription
config = {
"audio_url": FILE_URL,
"redact_pii":True,
"speaker_labels":True,
"format_text":True,
"punctuate":True,
"speech_model":"universal",
"language_detection":True
}
config["redact_pii_policies"] = [
"medical_condition",
"email_address",
"phone_number",
"banking_information",
"credit_card_number",
"credit_card_cvv",
"date_of_birth",
"person_name"
]
url = base_url + "/v2/transcript"
response = requests.post(url, json=config, headers=headers)
transcript_id = response.json()['id']
polling_endpoint = base_url + "/v2/transcript/" + transcript_id
while True:
transcription_result = requests.get(polling_endpoint, headers=headers).json()
transcription_text = transcription_result['text']
if transcription_result['status'] == 'completed':
print(f"Transcript Text:", transcription_text)
break
elif transcription_result['status'] == 'error':
raise RuntimeError(f"Transcription failed: {transcription_result['error']}")
else:
time.sleep(3)
import requests
import time
base_url = "https://api.assemblyai.com"
headers = {
# Replace with your chosen API key, this is the "default" account api key
"authorization": "<>"
}
# URL of the file to transcribe
FILE_URL = "https://cdn.assemblyai.com/upload/<>"
# You can set additional parameters for the transcription
config = {
"audio_url": FILE_URL,
"redact_pii":True,
"speaker_labels":True,
"format_text":True,
"punctuate":True,
"speech_model":"universal",
"language_detection":True
}
config["speech_understanding"] = {
"request": {
"speaker_identification": {
"speaker_type": "role",
"known_values": [
"Agent",
"Customer"
]
}
}
}
config["redact_pii_policies"] = [
"medical_condition",
"email_address",
"phone_number",
"banking_information",
"credit_card_number",
"credit_card_cvv",
"date_of_birth",
"person_name"
]
url = base_url + "/v2/transcript"
response = requests.post(url, json=config, headers=headers)
transcript_id = response.json()['id']
polling_endpoint = base_url + "/v2/transcript/" + transcript_id
while True:
transcription_result = requests.get(polling_endpoint, headers=headers).json()
transcription_text = transcription_result['text']
if transcription_result['status'] == 'completed':
print(f"Transcript Text:", transcription_text)
break
elif transcription_result['status'] == 'error':
raise RuntimeError(f"Transcription failed: {transcription_result['error']}")
else:
time.sleep(3)
But when I add speaker identification it brokes the PIIL
(data is fake, don't worry)
I apologize for putting the issue here, I didn't know where to put it
Metadata
Metadata
Assignees
Labels
No labels