update

Author: aboutphilippe

production_demo/README.md

@@ -63,11 +63,11 @@ And for each child workflow, for each step you can see how long the function sta
 - Python 3.10 or higher
 - Poetry (for dependency management)
 - Docker (for running the Restack services)
-- Local LLM provider (we use LMStudio and a Meta Llama 3.1 8B Instruct 4bit model in this example)
+- Local LLM provider (we use LMStudio and a Meta Llama 3.2 3B Instruct 4bit model in this example)
 
 ## Start LM stduio for local LLM provider
 
-Start local server with an open source model like llama-3.3-70b-instruct.
+Start local server with an open source model like llama-3.2-3b-instruct
 
 https://lmstudio.ai
 

production_demo/schedule_workflow.py

@@ -6,7 +6,7 @@ async def main():
 
     client = Restack()
 
-    workflow_id = f"{int(time.time() * 1000)}-ExampleWorkflow"
+    workflow_id = f"{int(time.time() * 1000)}-ChildWorkflow"
     run_id = await client.schedule_workflow(
         workflow_name="ChildWorkflow",
         workflow_id=workflow_id

production_demo/src/functions/evaluate.py

@@ -27,7 +27,7 @@ async def llm_evaluate(input: EvaluateInput) -> str:
 
     try:
         response = client.chat.completions.create(
-            model="llama-3.3-70b-instruct",
+            model="llama-3.2-3b-instruct",
             messages=[
                 {
                     "role": "user",

production_demo/src/functions/generate.py

@@ -17,7 +17,7 @@ async def llm_generate(input: GenerateInput) -> str:
 
     try:
         response = client.chat.completions.create(
-            model="llama-3.3-70b-instruct",
+            model="llama-3.2-3b-instruct",
             messages=[
                 {
                     "role": "user",

production_demo/src/workflows/child.py

@@ -8,7 +8,7 @@
     from src.functions.evaluate import llm_evaluate, EvaluateInput
 
 class ChildWorkflowInput(BaseModel):
-    name: str = Field(default='John Doe')
+    prompt: str = Field(default="Generate a random joke in max 20 words.")
 
 @workflow.defn()
 class ChildWorkflow:
@@ -21,7 +21,7 @@ async def run(self, input: ChildWorkflowInput):
 
         generated_text = await workflow.step(
             llm_generate,
-            GenerateInput(prompt="Generate a random joke in max 20 words."),
+            GenerateInput(prompt=input.prompt),
             task_queue="llm",
             start_to_close_timeout=timedelta(minutes=2)
         )
@@ -30,7 +30,7 @@ async def run(self, input: ChildWorkflowInput):
             llm_evaluate,
             EvaluateInput(generated_text=generated_text),
             task_queue="llm",
-            start_to_close_timeout=timedelta(minutes=2)
+            start_to_close_timeout=timedelta(minutes=5)
         )
 
         return {