Llama deploy and fastapi support (#23)

* -product pitch presentation

* -llama-deploy implementation, -implemented fastapi for all templates, -partial implementation of docker for templates

* implemented the llama-deploy and workflows with simplemq and rabbitmq

* - modified the docs

---------

Co-authored-by: pavanmantha <[email protected]>

Author: pavanjava

bootstraprag/cli.py

@@ -1,8 +1,6 @@
 import click
 import shutil
 from pathlib import Path
-import os
-import zipfile
 from InquirerPy import inquirer
 
 
@@ -36,7 +34,9 @@ def create(project_name, framework, template, observability):
             'rag-with-hyde',
             'rag-with-flare',
             'rag-with-self-correction',
-            'llama-agents-with-simpleq'
+            'llama-deploy-with-simplemq',
+            'llama-deploy-with-rabbitmq',
+            'llama-deploy-with-kafka'
         ]
     elif framework == 'None':
         framework = 'qdrant'

bootstraprag/templates/llamaindex/llama_agents_with_kafka/readme.md

@@ -19,3 +19,7 @@ WARN = 30
 INFO = 20
 DEBUG = 10
 NOTSET = 0
+
+# allowed values deploy_rag_workflow_with_retry_query_engine,
+# deploy_rag_workflow_with_retry_source_query_engine, deploy_rag_workflow_with_retry_guideline_query_engine
+ENABLED_WORKFLOW='deploy_rag_workflow_with_retry_source_query_engine'

bootstraprag/templates/llamaindex/llama_agents_with_rabbitmq/readme.md

@@ -0,0 +1,18 @@
+from llama_deploy import (
+    deploy_core,
+    ControlPlaneConfig,
+    SimpleMessageQueueConfig,
+)
+
+
+async def main():
+    await deploy_core(
+        control_plane_config=ControlPlaneConfig(),
+        message_queue_config=SimpleMessageQueueConfig(),
+    )
+
+
+if __name__ == "__main__":
+    import asyncio
+
+    asyncio.run(main())

bootstraprag/templates/llamaindex/llama_agents_with_simpleq/agents_core.py

@@ -0,0 +1,48 @@
+import os
+
+from llama_deploy import (
+    deploy_workflow,
+    WorkflowServiceConfig,
+    ControlPlaneConfig
+)
+from workflows.retry_query_engine_workflow import build_rag_workflow_with_retry_query_engine
+from dotenv import load_dotenv, find_dotenv
+
+
+_ = load_dotenv(find_dotenv())
+
+
+async def deploy_rag_workflow_with_retry_query_engine():
+    rag_workflow = build_rag_workflow_with_retry_query_engine()
+    try:
+        await deploy_workflow(
+            workflow=rag_workflow,
+            workflow_config=WorkflowServiceConfig(
+                host="127.0.0.1",
+                port=8002,
+                # service name matches the name of the workflow used in Agentic Workflow
+                service_name="rag_workflow_with_retry_query_engine",
+                description="RAG workflow",
+            ),
+            # Config controlled by env vars
+            control_plane_config=ControlPlaneConfig()
+        )
+    except Exception as e:
+        print(e)
+
+if __name__ == "__main__":
+    import asyncio
+    import nest_asyncio
+
+    nest_asyncio.apply()
+    try:
+        # deployment of workflow is driven by environmental variables
+        # if os.environ['ENABLED_WORKFLOW'] == 'deploy_rag_workflow_with_retry_query_engine':
+        #     asyncio.run(deploy_rag_workflow_with_retry_query_engine())
+        # elif os.environ['ENABLED_WORKFLOW'] == 'deploy_rag_workflow_with_retry_source_query_engine':
+        #     asyncio.run(deploy_rag_workflow_with_retry_source_query_engine())
+        # else:
+        #     asyncio.run(deploy_rag_workflow_with_retry_guideline_query_engine())
+        asyncio.run(deploy_rag_workflow_with_retry_query_engine())
+    except Exception as e:
+        print(e)

bootstraprag/templates/llamaindex/llama_agents_with_simpleq/main.py

@@ -0,0 +1,51 @@
+import os
+
+from llama_deploy import (
+    deploy_workflow,
+    WorkflowServiceConfig,
+    ControlPlaneConfig
+)
+from workflows.retry_query_engine_workflow import build_rag_workflow_with_retry_query_engine
+from workflows.retry_source_query_engine_workflow import build_rag_workflow_with_retry_source_query_engine
+from workflows.retry_guideline_query_engine_workflow import build_rag_workflow_with_retry_guideline_query_engine
+from dotenv import load_dotenv, find_dotenv
+
+
+_ = load_dotenv(find_dotenv())
+
+
+async def deploy_rag_workflow_with_retry_source_query_engine():
+    rag_workflow = build_rag_workflow_with_retry_source_query_engine()
+    try:
+        await deploy_workflow(
+            workflow=rag_workflow,
+            workflow_config=WorkflowServiceConfig(
+                host="127.0.0.1",
+                port=8003,
+                # service name matches the name of the workflow used in Agentic Workflow
+                service_name="rag_workflow_with_retry_source_query_engine",
+                description="RAG workflow",
+            ),
+            # Config controlled by env vars
+            control_plane_config=ControlPlaneConfig()
+        )
+    except Exception as e:
+        print(e)
+
+
+if __name__ == "__main__":
+    import asyncio
+    import nest_asyncio
+
+    nest_asyncio.apply()
+    try:
+        # deployment of workflow is driven by environmental variables
+        # if os.environ['ENABLED_WORKFLOW'] == 'deploy_rag_workflow_with_retry_query_engine':
+        #     asyncio.run(deploy_rag_workflow_with_retry_query_engine())
+        # elif os.environ['ENABLED_WORKFLOW'] == 'deploy_rag_workflow_with_retry_source_query_engine':
+        #     asyncio.run(deploy_rag_workflow_with_retry_source_query_engine())
+        # else:
+        #     asyncio.run(deploy_rag_workflow_with_retry_guideline_query_engine())
+        asyncio.run(deploy_rag_workflow_with_retry_source_query_engine())
+    except Exception as e:
+        print(e)