feat(example): add an langchain-based chatbot example, and fix a bug

.changeset/brave-countries-build.md

@@ -0,0 +1,5 @@
+---
+"@plutolang/pluto-infra": patch
+---
+
+enhance(sdk): remove the fuzzyArn, instead use the lazy value from pulumi.

.changeset/breezy-cats-mate.md

@@ -0,0 +1,5 @@
+---
+"@plutolang/pluto": patch
+---
+
+feat(sdk): export the table name and partition key of the created dynamodb instance to users

.changeset/config.json

@@ -15,6 +15,8 @@
     "quickstart",
     "gpt2-hf-sagemaker",
     "langchain-llama2-sagemaker",
-    "quickstart-python"
+    "quickstart-python",
+    "langchain-llama2-chatbot-sagemaker",
+    "langchain-llama2-chatbot-sagemaker-python"
   ]
 }

.changeset/tough-swans-behave.md

@@ -0,0 +1,5 @@
+---
+"@plutolang/static-deducer": patch
+---
+
+fix(deducer): incorrectly generate duplicate client statements when one code section accesses a single resource multiple times

.github/workflows/release.yml

@@ -29,6 +29,10 @@ jobs:
         with:
           node-version: 20.x
 
+      - uses: actions/setup-python@v5
+        with:
+          python-version: "3.10"
+
       - name: Setup PNPM
         uses: pnpm/action-setup@v2
         with:
@@ -38,7 +42,7 @@ jobs:
         run: bash scripts/prepare.sh
 
       - name: Install Dependencies
-        run: pnpm install
+        run: pnpm install && pip install -r requirements.txt
 
       - name: Creating .npmrc
         run: |

components/deducers/static/src/closure/store-closure.ts

@@ -47,8 +47,14 @@ function generateSourceCode(
 ): string {
   let cirCode = imports + "\n";
 
+  const resources = new Set<string>();
   // Find the dependencies of this CIR and build corresponding instances.
   for (const dependentRes of dependentResources) {
+    if (resources.has(dependentRes.name)) {
+      continue;
+    }
+    resources.add(dependentRes.name);
+
     // TODO: verify if the buildClient function exists. If it does not, use the original statement.
     cirCode += dependentRes.imports + "\n";
     cirCode += `const ${dependentRes.name} = ${dependentRes.type}.buildClient(${dependentRes.parameters});\n`;

examples/langchain-llama2-chatbot-sagemaker-python/.gitignore

@@ -0,0 +1,38 @@
+# See https://help.github.com/articles/ignoring-files/ for more about ignoring files.
+
+# dependencies
+node_modules
+.pnp
+.pnp.js
+
+# testing
+coverage
+
+# misc
+.DS_Store
+*.pem
+
+# debug
+npm-debug.log*
+yarn-debug.log*
+yarn-error.log*
+
+# local env files
+.env
+.env.local
+.env.development.local
+.env.test.local
+.env.production.local
+
+# turbo
+.turbo
+
+# vercel
+.vercel
+
+# compilation result
+**/dist/
+
+# pluto
+.pluto/**/state/
+.pluto/*.test.*/

examples/langchain-py/.pluto/pluto.yml → examples/langchain-llama2-chatbot-sagemaker-python/.pluto/pluto.yml

@@ -2,6 +2,6 @@ current: aws
 language: python
 stacks:
   - configs: {}
-    provisionType: Pulumi
     name: aws
-    platformType: AWS
+    platformType: AWS
+    provisionType: Pulumi

examples/langchain-llama2-chatbot-sagemaker-python/README.md

@@ -0,0 +1,6 @@
+# Building a Llama2 Conversational Chatbot with AWS and LangChain
+
+The difference between this and the [TypeScript version of the Pluto example](../langchain-llama2-chatbot-sagemaker/) lies in its implementation in Python, which has not yet been successfully deployed. The issues encountered include:
+
+1. When deploying across platforms, dependencies such as Numpy and Pydantic cannot be correctly installed on the target platform.
+2. The size of the compressed package exceeds the AWS Lambda limit of 50MB.

examples/langchain-llama2-chatbot-sagemaker-python/README_zh.md

@@ -0,0 +1,6 @@
+# 基于 AWS 和 LangChain 的 Llama2 会话聊天机器人
+
+与 [TypeScript 版本的 Pluto 示例](../langchain-llama2-chatbot-sagemaker/)应用区别在于使用 Python 实现，尚未成功部署，存在的问题包括：
+
+1. 跨平台部署时，Numpy、Pydantic 等依赖包无法在目标平台正确安装。
+2. 压缩包大小超过 AWS Lambda 限制，50MB。

examples/langchain-llama2-chatbot-sagemaker-python/app/main.py

@@ -0,0 +1,127 @@
+import os
+import json
+from typing import Dict
+
+from pluto_client.sagemaker import SageMaker, SageMakerOptions
+from pluto_client import Router, HttpRequest, HttpResponse, KVStore
+
+from langchain.prompts import PromptTemplate
+from langchain.memory.buffer import ConversationBufferMemory
+from langchain.chains.conversation.base import ConversationChain
+from langchain_community.chat_message_histories.dynamodb import (
+    DynamoDBChatMessageHistory,
+)
+from langchain_community.llms.sagemaker_endpoint import (
+    SagemakerEndpoint,
+    LLMContentHandler,
+)
+
+"""
+Deploy the Llama2 model on AWS SageMaker using the Hugging Face Text Generation Inference (TGI)
+container. If you're unable to deploy the model because of the instance type, consider using the 
+TinyLlama-1.1B-Chat-v1.0 model, which is compatible with the ml.m5.xlarge instance.
+
+Below is a set up minimum requirements for each model size of Llama2 model:
+```
+Model      Instance Type    Quantization    # of GPUs per replica
+Llama 7B   ml.g5.2xlarge    -               1
+Llama 13B  ml.g5.12xlarge   -               4
+Llama 70B  ml.g5.48xlarge   bitsandbytes    8
+Llama 70B  ml.p4d.24xlarge  -               8
+```
+
+The initial limit set for these instances is zero. If you need more, you can request an increase
+in quota via the [AWS Management Console](https://console.aws.amazon.com/servicequotas/home).
+"""
+sagemaker = SageMaker(
+    "llama2",
+    "763104351884.dkr.ecr.us-east-1.amazonaws.com/huggingface-pytorch-tgi-inference:2.1.1-tgi1.4.0-gpu-py310-cu121-ubuntu20.04",
+    SageMakerOptions(
+        instanceType="ml.g5.2xlarge",
+        envs={
+            "HF_MODEL_ID": "meta-llama/Llama-2-7b-chat-hf",
+            "HF_TASK": "text-generation",
+            # If you want to deploy the Meta Llama2 model, you need to request a permission and
+            # prepare the token. You can get the token from https://huggingface.co/settings/tokens
+            "HUGGING_FACE_HUB_TOKEN": "hf_EmXPwpnyHoNrxxxxxxxxx",
+        },
+    ),
+)
+
+router = Router("chatbot")
+conversations = KVStore("conversations")
+
+
+class ContentHandler(LLMContentHandler):
+    content_type = "application/json"
+    accepts = "application/json"
+
+    def transform_input(self, prompt: str, model_kwargs: Dict) -> bytes:
+        input_str = json.dumps({"inputs": prompt, "parameters": model_kwargs})
+        return input_str.encode("utf-8")
+
+    def transform_output(self, output: bytes) -> str:
+        response_json = json.loads(output.decode("utf-8"))
+        return response_json[0]["generated_text"]
+
+
+def get_aws_region() -> str:
+    aws_region = os.environ.get("AWS_REGION")
+    if aws_region is None:
+        raise ValueError("AWS_REGION environment variable must be set")
+    return aws_region
+
+
+llm = SagemakerEndpoint(
+    endpoint_name=sagemaker.endpoint_name,  # SageMaker endpoint name
+    region_name=get_aws_region(),
+    content_handler=ContentHandler(),
+)
+
+
+def chat_handler(req: HttpRequest) -> HttpResponse:
+    query = req.query["query"]
+    sessionid = req.query["sessionid"]
+    if isinstance(sessionid, list):
+        sessionid = sessionid[0]
+
+    memory = ConversationBufferMemory(
+        chat_memory=DynamoDBChatMessageHistory(
+            table_name=conversations.aws_table_name,  # DynamoDB table name
+            primary_key_name=conversations.aws_partition_key,  # DynamoDB partition key
+            session_id=sessionid,
+        )
+    )
+
+    promptTemplate = PromptTemplate.from_template(
+        """<|system|>
+You are a cool and aloof robot, answering questions very briefly and directly.
+
+Context:
+{history}</s>
+<|user|>
+{query}</s>
+<|assistant|>"""
+    )
+
+    chain = ConversationChain(
+        llm=llm,
+        memory=memory,
+        prompt=promptTemplate,
+    )
+
+    result = chain({"query": query})
+    print(result)
+
+    return HttpResponse(status_code=200, body="hello world")
+
+    # TODO: Use the following statement to help the deducer identify the right relationship between
+    # Lambda and other resources. This will be used to grant permission for the Lambda instance to
+    # access the SageMaker endpoint and the DynamoDB. This code should be removed after the deducer
+    # supports the analysis of libraries.
+    conversations.get("")
+    conversations.set("", "")
+    sagemaker.invoke("")
+
+
+router.get("/chatbot", chat_handler)

examples/langchain-llama2-chatbot-sagemaker-python/package.json

@@ -0,0 +1,20 @@
+{
+  "name": "langchain-llama2-chatbot-sagemaker-python",
+  "private": true,
+  "version": "0.0.1",
+  "scripts": {
+    "test:dev": "pluto test --sim",
+    "test:prod": "pluto test",
+    "deploy": "pluto deploy",
+    "destroy": "pluto destroy"
+  },
+  "dependencies": {},
+  "devDependencies": {
+    "@types/node": "^20",
+    "typescript": "^5.2.2",
+    "@plutolang/base": "latest",
+    "@plutolang/pluto-infra": "latest",
+    "@pulumi/pulumi": "^3.88.0"
+  },
+  "main": "dist/index.js"
+}

examples/langchain-llama2-chatbot-sagemaker-python/requirements.txt

@@ -0,0 +1 @@
+pluto_client

examples/langchain-llama2-chatbot-sagemaker/.gitignore

@@ -0,0 +1,38 @@
+# See https://help.github.com/articles/ignoring-files/ for more about ignoring files.
+
+# dependencies
+node_modules
+.pnp
+.pnp.js
+
+# testing
+coverage
+
+# misc
+.DS_Store
+*.pem
+
+# debug
+npm-debug.log*
+yarn-debug.log*
+yarn-error.log*
+
+# local env files
+.env
+.env.local
+.env.development.local
+.env.test.local
+.env.production.local
+
+# turbo
+.turbo
+
+# vercel
+.vercel
+
+# compilation result
+**/dist/
+
+# pluto
+.pluto/**/state/
+.pluto/*.test.*/

examples/langchain-llama2-chatbot-sagemaker/.pluto/pluto.yml

@@ -0,0 +1,7 @@
+current: dev
+language: typescript
+stacks:
+  - configs: {}
+    name: dev
+    platformType: AWS
+    provisionType: Pulumi