Merge pull request #27 from slashml/azeem-library-refactor

refactor library and examples

Author: eff-kay

examples/speech_summarized.py

@@ -1,58 +1,69 @@
-
-from slashml import SpeechToText, Summarization
+from slashml import SpeechToText, TextSummarization
 import time
 
-# Initialize your model
-model = SpeechToText()
 
-service_provider = model.ServiceProvider.AWS
+def speech_to_text(uploaded_audio_url, service_provider, api_key):
+    # Initialize model
+    model = SpeechToText(api_key=api_key)
+
+    # Submit transcription request
+    job = model.transcribe(uploaded_audio_url, service_provider=service_provider)
+
+    assert job.status != "ERROR", f"{job}"
+    print(f"Job ID: {job.id}")
+
+    # check job status
+    response = model.status(job.id, service_provider=service_provider)
+
+    while response.status == "IN_PROGRESS":
+        time.sleep(30)
+        response = model.status(job.id, service_provider=service_provider)
+        print(f"Response = {response}. Retrying in 30 seconds")
+
+    return response
+
 
-# long audio file already uploaded
-upload_url = 'https://slashml.s3.ca-central-1.amazonaws.com/fda70f6a-6057-4541-adf1-2cf4f4182929'
+def summarize(text, service_provider, api_key):
+    # Initialize model
+    model = TextSummarization(api_key=api_key)
 
-# # choose a service provider and transcribe
-print('transcribing file')
-job = model.transcribe(
-    upload_url, 
-    service_provider=service_provider
-    )
+    # Submit request
+    job = model.summarize(text=text, service_provider=service_provider)
 
-assert job.status != 'ERROR', f'{job}'
+    assert job.status != "ERROR", f"{job}"
+    print(f"Job ID: {job.id}")
 
-print('transcription job id', job.id)
+    # Check job status
+    response = model.status(job.id, service_provider=service_provider)
 
-# check the status
-response = model.status(job.id, service_provider=service_provider)
+    # Keep checking job status until the task is complete
+    while response.status == "PENDING":
+        print(f"Response = {response}. Retrying in 30 seconds")
+        time.sleep(30)
+        response = model.status(job.id, service_provider=service_provider)
 
-while response.status != 'COMPLETED':
-    print(f'response got', response , 'trying again')
-    time.sleep(30)
-    response = model.status(job.id, service_provider = service_provider)
+    return response
 
-text_to_summarize = response.transcription_data.transcription
 
-print('summarizing text', text_to_summarize)
+# Replace `API_KEY` with your SlasML API token. This example still runs without
+# the API token but usage will be limited
+API_KEY = "b503d137d229fdd4085f59e9ea06ac95d2706182"
 
-model:Summarization = Summarization()
-service_provider = model.ServiceProvider.OPENAI
+service_provider_speech_to_text = SpeechToText.ServiceProvider.AWS
+service_provider_summarize = TextSummarization.ServiceProvider.OPENAI
 
-print('summarizing text')
-# choose a service provider and transcribe
-job = model.summarize(
-    text=text_to_summarize, 
-    service_provider=service_provider
-    )
+# 10 minute audio file already uploaded
+uploaded_url = (
+    "https://slashml.s3.ca-central-1.amazonaws.com/fda70f6a-6057-4541-adf1-2cf4f4182929"
+)
 
-assert job.status != 'ERROR', f'{job}'
 
-print('summarization job id', job.id)
+response_speech_to_text = speech_to_text(uploaded_url, service_provider_speech_to_text, API_KEY)
 
-# check the status
-response = model.status(job.id, service_provider=service_provider)
+transcribed_text = response_speech_to_text.transcription_data.transcription
+print (f"Transcribed Text = {transcribed_text}")
 
-while response.status != 'COMPLETED':
-    print(f'response got', response , 'trying again')
-    time.sleep(30)
-    response = model.status(job.id, service_provider = service_provider)
+response_summarize = summarize(transcribed_text, service_provider_summarize, API_KEY)
+summary = response_summarize.summarization_data
 
-print(response.summarization_data)
+print (f"Summarized Text = {summary}")

examples/speech_to_text.py

@@ -1,30 +1,46 @@
-from slashml.slashml import SpeechToText
+from slashml import SpeechToText
 import time
 
-# Initialize your model
-model = SpeechToText()
-# model = SpeechToText()
 
-service_provider = model.ServiceProvider.ASSEMBLY
+def speech_to_text(audio_filepath, service_provider, api_key):
+    # Initialize model
+    model = SpeechToText(api_key=api_key)
 
-uploaded_file = model.upload_audio("test.mp3")
-print(uploaded_file)
+    # Upload audio
+    uploaded_file = model.upload_audio(audio_filepath)
+    print(f"file uploaded: {uploaded_file}")
 
-# # choose a service provider and transcribe
-job = model.transcribe(
-    uploaded_file['upload_url'], 
-    service_provider=service_provider
+    # Submit transcription request
+    job = model.transcribe(
+        uploaded_file["upload_url"], service_provider=service_provider
     )
 
-assert job.status != 'ERROR', f'{job}'
-print('transcription job id', job.id)
+    assert job.status != "ERROR", f"{job}"
+    print(f"Job ID: {job.id}")
 
-# check the status
-response = model.status(job.id, service_provider=service_provider)
+    # check job status
+    response = model.status(job.id, service_provider=service_provider)
 
-while response.status != 'COMPLETED':
-    time.sleep(30)
-    response = model.status(job.id, service_provider = service_provider)
-    print(f'response got', response , 'trying agagin')
+    while response.status == "IN_PROGRESS":
+        print(f"Response = {response}. Retrying in 30 seconds")
+        time.sleep(30)
+        response = model.status(job.id, service_provider=service_provider)
 
-print(response.transcription_data.transcription)
+    return response
+
+
+# Replace `API_KEY` with your SlasML API token. This example still runs without
+# the API token but usage will be limited
+API_KEY = None
+service_provider = SpeechToText.ServiceProvider.ASSEMBLY
+audio_filepath = "test.mp3"
+
+# Find all the service providers that we support by running the choices() method
+print(f"Available providers: {SpeechToText.ServiceProvider.choices()}")
+print(f"Selected provider: {service_provider}")
+
+
+response = speech_to_text(
+    audio_filepath=audio_filepath, service_provider=service_provider, api_key=API_KEY
+)
+print(f"{response}\n\nTranscription = {response.transcription_data.transcription}")

examples/summarize.py

@@ -1,30 +1,37 @@
-
-from slashml import Summarization
+from slashml import TextSummarization
 import time
 
-# Initialize your model
-model = Summarization()
-# model = Summarization()
-service_provider = model.ServiceProvider.OPENAI
+def summarize(text, service_provider, api_key):
+    # Initialize model
+    model = TextSummarization(api_key=api_key)
+
+    # Submit request
+    job = model.summarize(text=text, service_provider=service_provider)
+
+    assert job.status != "ERROR", f"{job}"
+    print(f"Job ID: {job.id}")
 
-in_text = '''A good writer doesn't just think, and then write down what he thought, as a sort of transcript. A good writer will almost always discover new things in the process of writing. And there is, as far as I know, no substitute for this kind of discovery. Talking about your ideas with other people is a good way to develop them. But even after doing this, you'll find you still discover new things when you sit down to write. There is a kind of thinking that can only be done by writing.'''
+    # Check job status
+    response = model.status(job.id, service_provider=service_provider)
 
-# # choose a service provider and transcribe
-job = model.summarize(
-    text=in_text,
-    service_provider=service_provider
-    )
+    # Keep checking job status until the task is complete
+    while response.status == "PENDING":
+        print(f"Response = {response}. Retrying in 30 seconds")
+        time.sleep(30)
+        response = model.status(job.id, service_provider=service_provider)
 
-assert job.status != 'ERROR', f'{job}'
+    return response
 
-print('transcription job id', job.id)
 
-# check the status
-response = model.status(job.id, service_provider = service_provider)
+# Replace `API_KEY` with your SlasML API token. This example still runs without
+# the API token but usage will be limited
+API_KEY = "b503d137d229fdd4085f59e9ea06ac95d2706182"
+service_provider = TextSummarization.ServiceProvider.OPENAI
+input_text = """A good writer doesn't just think, and then write down what he thought, as a sort of transcript. A good writer will almost always discover new things in the process of writing. And there is, as far as I know, no substitute for this kind of discovery. Talking about your ideas with other people is a good way to develop them. But even after doing this, you'll find you still discover new things when you sit down to write. There is a kind of thinking that can only be done by writing."""
 
-while response.status != 'COMPLETED':
-    time.sleep(30)
-    response = model.status(job['id'], service_provider = service_provider)
-    print(f'response got', response , 'trying again')
+# Find all the service providers that we support by running the choices() method
+print(f"Available providers: {TextSummarization.ServiceProvider.choices()}")
+print(f"Selected provider: {service_provider}")
 
-print(response)
+response = summarize(text=input_text, service_provider=service_provider, api_key=API_KEY)
+print(f"{response}\n\nSummary = {response.summarization_data}")

examples/text_to_speech.py

@@ -1,16 +1,19 @@
-
 from slashml import TextToSpeech
 
-# Initialize your model
-model = TextToSpeech()
-# model = SpeechToText()
+# Replace `API_KEY` with your SlasML API token. This example still runs without
+# the API token but usage will be limited
+API_KEY = None
+service_provider = TextToSpeech.ServiceProvider.AWS
+input_text = "To be or not to be, that is the question!"
+
+# Find all the service providers that we support by running the choices() method
+print(f"Available providers: {TextToSpeech.ServiceProvider.choices()}")
+print(f"Selected provider: {service_provider}")
 
-service_provider = model.ServiceProvider.AWS
+model = TextToSpeech(api_key=API_KEY)
 
-# # choose a service provider and transcribe
-job = model.speechify(
-    text ='this is my text',
-    service_provider=service_provider
-    )
+# Submit request
+job = model.speechify(text=input_text, service_provider=service_provider)
+print(job)
 
-print(job)
+print (f"\nYou can access the audio file here: {job.audio_url}")

requires-install.txt

@@ -1 +1,2 @@
 requests==2.28.1
+addict==2.4.0

slashml/__init__.py

@@ -1,2 +1,4 @@
-from .slashml import SpeechToText, Summarization  # noqa: F401,E402
+from .text_summarization import TextSummarization  # noqa: F401,E402
+from .speech_to_text import SpeechToText  # noqa: F401,E402
 from .text_to_speech import TextToSpeech  # noqa: F401,E402
+