Merge pull request #137 from restackio/audioEndpoint

Update audio translation example

Author: aboutphilippe

audio_transcript/README.md

@@ -1,71 +1,65 @@
-# Restack AI - Audio transcript and translation example
-
-This example showcases how to transcribe an mp3 audio and then later translate the generated text to a target language, all done in a single workflow defined with Restack AI.
+# Restack AI - Audio translation example
 
+This example showcases how to transcribe an mp3 audio and then translate the generated text to a target language, all done in a single workflow defined with Restack AI.
 
 ## Prerequisites
 
+- Docker (for running Restack)
 - Python 3.10 or higher
 - Poetry (for dependency management)
-- Docker (for running the Restack services)
 
-## Usage
+## Start Restack
+
+To start the Restack, use the following Docker command:
+
+```bash
+docker run -d --pull always --name restack -p 5233:5233 -p 6233:6233 -p 7233:7233 ghcr.io/restackio/restack:main
+```
+
+## Start python shell
 
-1. Run Restack local engine with Docker:
+```bash
+poetry env use 3.10 && poetry shell
+```
 
-   ```bash
-   docker run -d --pull always --name restack -p 5233:5233 -p 6233:6233 -p 7233:7233 ghcr.io/restackio/restack:main
-   ```
+## Install dependencies
 
-2. Open the web UI to see the workflows:
+```bash
+poetry install
+```
 
-   ```bash
-   http://localhost:5233
-   ```
+```bash
+poetry env info # Optional: copy the interpreter path to use in your IDE (e.g. Cursor, VSCode, etc.)
+```
 
-3. Clone this repository:
+```bash
+poetry run dev
+```
 
-   ```bash
-   git clone https://github.com/restackio/examples-python
-   cd examples-python/examples/get-started
-   ```
-  
-4. Create .env file with: STRIPE_SECRET_KEY and OPENAI_API_KEY
+## Run workflows
 
-4. Install dependencies using Poetry:
+### from UI
 
-   ```bash
-   poetry env use 3.12
-   ```
+You can run workflows from the UI by clicking the "Run" button.
 
-   ```bash
-   poetry shell
-   ```
+![Run workflows from UI](./ui-screenshot.png)
 
-   ```bash
-   poetry install
-   ```
+### from API
 
-   ```bash
-   poetry env info # Optional: copy the interpreter path to use in your IDE (e.g. Cursor, VSCode, etc.)
-   ```
+You can run workflows from the API by using the generated endpoint:
 
-5. Run the services:
+`POST http://localhost:6233/api/workflows/TranscribeTranslateWorkflow`
 
-   ```bash
-   poetry run services
-   ```
+### from any client
 
-   This will start the Restack service with the defined workflows and functions.
+You can run workflows with any client connected to Restack, for example:
 
-6. In a new terminal, schedule the workflow:
+```bash
+poetry run schedule
+```
 
-   ```bash
-   poetry shell
-   ```
+executes `schedule_workflow.py` which will connect to Restack and execute the `TranscribeTranslateWorkflow` workflow.
 
-   ```bash
-   poetry run schedule
-   ```
+## Deploy on Restack Cloud
 
-   This will schedule the `TranscribeTranslateWorkflow` and print the result.
+To deploy the application on Restack, you can create an account at [https://console.restack.io](https://console.restack.io)

audio_transcript/pyproject.toml

@@ -1,7 +1,7 @@
 [tool.poetry]
 name = "audio_transcript"
 version = "0.0.1"
-description = "Send emails with sendgrid"
+description = "Transcribe audio with OpenAI Whisper and translate the text with OpenAI GPT-4o-mini"
 authors = [
     "Restack Team <[email protected]>",
 ]
@@ -12,14 +12,16 @@ packages = [{include = "src"}]
 python = ">=3.10,<4.0"
 pydantic = "^2.10.3"
 python-dotenv = "1.0.1"
-openai = "^1.57.2"
-restack-ai = "^0.0.48"
+restack-ai = "^0.0.53"
+openai = "^1.59.8"
+watchfiles = "^1.0.4"
 
 [build-system]
 requires = ["poetry-core"]
 build-backend = "poetry.core.masonry.api"
 
 [tool.poetry.scripts]
+dev = "src.services:watch_services"
 services = "src.services:run_services"
 schedule = "schedule_workflow:run_schedule_workflow"
 schedule_failure = "schedule_workflow_failure:run_schedule_workflow_failure"

audio_transcript/schedule_workflow.py

@@ -1,29 +1,23 @@
 import asyncio
 import time
 from restack_ai import Restack
-from dataclasses import dataclass
-
 from dotenv import load_dotenv
+from src.workflows.transcribe_translate import WorkflowInputParams
 
 load_dotenv()
 
-@dataclass
-class InputParams:
-    file_path: str
-    target_language: str
-
-async def main(input: InputParams):
+async def main(input: WorkflowInputParams):
     client = Restack()
 
     workflow_id = f"{int(time.time() * 1000)}-TranscribeTranslateWorkflow"
 
     run_id = await client.schedule_workflow(
         workflow_name="TranscribeTranslateWorkflow",
         workflow_id=workflow_id,
-        input={
-            "file_path": input.file_path,
-            "target_language": input.target_language
-        }
+        input=WorkflowInputParams(
+            file_path=input.file_path,
+            target_language=input.target_language
+        )
     )
 
     await client.get_workflow_result(
@@ -34,7 +28,7 @@ async def main(input: InputParams):
     exit(0)
 
 def run_schedule_workflow():
-    asyncio.run(main(InputParams(file_path="./test.mp3", target_language="Spanish")))
+    asyncio.run(main(WorkflowInputParams(file_path="./test.mp3", target_language="Spanish")))
 
 if __name__ == "__main__":
     run_schedule_workflow()

audio_transcript/src/functions/transcribe_audio.py

@@ -1,4 +1,4 @@
-from restack_ai.function import function, FunctionFailure
+from restack_ai.function import function, FunctionFailure, log
 from dataclasses import dataclass
 from openai import OpenAI
 import os
@@ -17,10 +17,13 @@ async def transcribe_audio(input: TranscribeAudioInput):
 
     client = OpenAI(api_key=os.environ.get("OPENAI_API_KEY"))
 
-    response = client.audio.transcriptions.create(
-        model="whisper-1",
-        file=open(input.file_path, "rb")
-    )
+    try:
+      response = client.audio.transcriptions.create(
+          model="whisper-1",
+          file=open(input.file_path, "rb")
+      )
+    except Exception as error:
+      log.error("An error occurred during transcription", error)
 
     return response.text
 

audio_transcript/src/functions/translate_text.py

@@ -18,19 +18,22 @@ async def translate_text(input: TranslateTextInput):
 
     client = OpenAI(api_key=os.environ.get("OPENAI_API_KEY"))
 
-    response = client.chat.completions.create(
-        model="gpt-4o-mini",
-        messages=[
-            {
-                "role": "system",
-                "content": "You are a helpful assistant that translates text from one language to another."
-            },
-            {
-                "role": "user",
-                "content": f"Translate the following text to {input.target_language}: {input.text}"
-            }
-        ]
-    )
+    try:
+      response = client.chat.completions.create(
+          model="gpt-4o-mini",
+          messages=[
+              {
+                  "role": "system",
+                  "content": "You are a helpful assistant that translates text from one language to another."
+              },
+              {
+                  "role": "user",
+                  "content": f"Translate the following text to {input.target_language}: {input.text}"
+              }
+          ]
+      )
+    except Exception as error:
+      log.error("An error occurred during translation", error)
 
     return response.choices[0].message.content
 

audio_transcript/src/services.py

@@ -3,6 +3,9 @@
 from src.workflows.transcribe_translate import TranscribeTranslateWorkflow
 from src.functions.transcribe_audio import transcribe_audio
 from src.functions.translate_text import translate_text
+from watchfiles import run_process
+import webbrowser
+import os
 
 async def main():
     await asyncio.gather(
@@ -13,7 +16,16 @@ async def main():
     )
 
 def run_services():
-    asyncio.run(main())
+    try:
+        asyncio.run(main())
+    except KeyboardInterrupt:
+        print("Service interrupted by user. Exiting gracefully.")
+
+def watch_services():
+    watch_path = os.getcwd()
+    print(f"Watching {watch_path} and its subdirectories for changes...")
+    webbrowser.open("http://localhost:5233")
+    run_process(watch_path, recursive=True, target=run_services)
 
 if __name__ == "__main__":
-    run_services()
+       run_services()

audio_transcript/src/workflows/transcribe_translate.py

@@ -1,14 +1,12 @@
 from restack_ai.workflow import workflow, import_functions, log
-from dataclasses import dataclass
-
+from pydantic import BaseModel, Field
 with import_functions():
     from src.functions.transcribe_audio import transcribe_audio, TranscribeAudioInput
     from src.functions.translate_text import translate_text, TranslateTextInput
 
-@dataclass
-class WorkflowInputParams:
-    file_path: str
-    target_language: str
+class WorkflowInputParams(BaseModel):
+    file_path: str = Field(default="/test.mp3")
+    target_language: str = Field(default="fr")
 
 @workflow.defn()
 class TranscribeTranslateWorkflow:

audio_transcript/ui-screenshot.png

@@ -26,7 +26,6 @@ A Restack backend application should be structured as follows:
     - env.example
     - README.md
     - Dockerfile
-    - restack_up.py
 
 All these files are mandatory.