feat(awel): New MessageConverter and more AWEL operators

[fangyinc]

Description

1. New MessageConverter to support unified model message adaptation.
2. Model layer multi-version message, the previous one was v1, the new version in the sdk defaults to v2
   • v1 is previous usage.
   • v2 does not automatically convert messages, you need to use MessageConverter
3. New ChatPromptTemplate
4. More built-in AWEL operators.

How Has This Been Tested?

1. make test
2. Run examples in examples/sdk and examples/awel

Snapshots:

Include snapshots for easier review.

Checklist:

• My code follows the style guidelines of this project
• I have already rebased the commits and make the commit message conform to the project standard.
• I have performed a self-review of my own code
• I have commented my code, particularly in hard-to-understand areas
• I have made corresponding changes to the documentation
• Any dependent changes have been merged and published in downstream modules

[fangyinc]

If you run AWEL file with dev mode, it will show AWEL DAG tree and graph, eg.

python examples/awel/data_analyst_assistant.py

DAG Tree:

HttpTrigger(node_id=1be1aa9a-0ae1-47db-ac44-818a31242d55)
 -> PromptTemplateBuilderOperator(node_id=b4156cd3-ca2f-4855-881d-92c194b605f9)
|  -> HistoryDynamicPromptBuilderOperator(node_id=50bcf0c0-3da4-4486-8b71-05e44ab680ad)
|    -> JoinOperator(node_id=61286112-1c4b-4b80-81ec-4fe662f901e3)
|      -> LLMBranchOperator(node_id=0489a3e3-4466-4696-88a6-407d512d00fa)
|        -> LLMOperator(node_id=ec336bd2-2a61-46e1-a56d-7609c2e1b725, node_name=llm_task)
|       |  -> MapOperator(node_id=6a15ac90-43d5-4037-a0c9-0de287e9eb86)
|       |    -> JoinOperator(node_id=edf31b66-c620-473a-94af-a8f8732fcb06)
|        -> StreamingLLMOperator(node_id=f050449a-a04b-4c76-b178-1e0aff787459, node_name=streaming_llm_task)
|          -> OpenAIStreamingOutputOperator(node_id=6c327808-ba77-4bf2-9720-69a55155d1d3)
|            -> JoinOperator(node_id=edf31b66-c620-473a-94af-a8f8732fcb06)
 -> MapOperator(node_id=20425277-ef51-4bd7-bd1f-684b4490b699)
|  -> ServePreChatHistoryLoadOperator(node_id=e1f647d7-48e6-4448-b8ec-ae05fc04d969)
|    -> BufferedConversationMapperOperator(node_id=fa961d16-4eec-4d98-a7fe-7845988f912a)
|      -> HistoryDynamicPromptBuilderOperator(node_id=50bcf0c0-3da4-4486-8b71-05e44ab680ad)
|        -> JoinOperator(node_id=61286112-1c4b-4b80-81ec-4fe662f901e3)
|          -> LLMBranchOperator(node_id=0489a3e3-4466-4696-88a6-407d512d00fa)
|            -> LLMOperator(node_id=ec336bd2-2a61-46e1-a56d-7609c2e1b725, node_name=llm_task)
|           |  -> MapOperator(node_id=6a15ac90-43d5-4037-a0c9-0de287e9eb86)
|           |    -> JoinOperator(node_id=edf31b66-c620-473a-94af-a8f8732fcb06)
|            -> StreamingLLMOperator(node_id=f050449a-a04b-4c76-b178-1e0aff787459, node_name=streaming_llm_task)
|              -> OpenAIStreamingOutputOperator(node_id=6c327808-ba77-4bf2-9720-69a55155d1d3)
|                -> JoinOperator(node_id=edf31b66-c620-473a-94af-a8f8732fcb06)
 -> MapOperator(node_id=f9019745-6093-41d6-ab5a-97d788645c2a)
|  -> HistoryDynamicPromptBuilderOperator(node_id=50bcf0c0-3da4-4486-8b71-05e44ab680ad)
|    -> JoinOperator(node_id=61286112-1c4b-4b80-81ec-4fe662f901e3)
|      -> LLMBranchOperator(node_id=0489a3e3-4466-4696-88a6-407d512d00fa)
|        -> LLMOperator(node_id=ec336bd2-2a61-46e1-a56d-7609c2e1b725, node_name=llm_task)
|       |  -> MapOperator(node_id=6a15ac90-43d5-4037-a0c9-0de287e9eb86)
|       |    -> JoinOperator(node_id=edf31b66-c620-473a-94af-a8f8732fcb06)
|        -> StreamingLLMOperator(node_id=f050449a-a04b-4c76-b178-1e0aff787459, node_name=streaming_llm_task)
|          -> OpenAIStreamingOutputOperator(node_id=6c327808-ba77-4bf2-9720-69a55155d1d3)
|            -> JoinOperator(node_id=edf31b66-c620-473a-94af-a8f8732fcb06)
 -> JoinOperator(node_id=61286112-1c4b-4b80-81ec-4fe662f901e3)
   -> LLMBranchOperator(node_id=0489a3e3-4466-4696-88a6-407d512d00fa)
     -> LLMOperator(node_id=ec336bd2-2a61-46e1-a56d-7609c2e1b725, node_name=llm_task)
    |  -> MapOperator(node_id=6a15ac90-43d5-4037-a0c9-0de287e9eb86)
    |    -> JoinOperator(node_id=edf31b66-c620-473a-94af-a8f8732fcb06)
     -> StreamingLLMOperator(node_id=f050449a-a04b-4c76-b178-1e0aff787459, node_name=streaming_llm_task)
       -> OpenAIStreamingOutputOperator(node_id=6c327808-ba77-4bf2-9720-69a55155d1d3)
         -> JoinOperator(node_id=edf31b66-c620-473a-94af-a8f8732fcb06)

DAG graph

Other example

python examples/awel/simple_chat_history_example.py

DAG Tree

HttpTrigger(node_id=b71f0ccc-4539-445e-8d87-1ec5ef6e7e83)
 -> MapOperator(node_id=23d11ff8-a1d4-4541-a1ff-adfb0eadd614)
   -> ChatHistoryPromptComposerOperator(node_id=3ee4c2af-d625-445b-a0f3-60863404d82e)
     -> LLMBranchOperator(node_id=bdc20002-b9e9-4ce8-9211-5aba83f4c16e)
       -> LLMOperator(node_id=b4c3fe48-a9d8-4968-a128-df5a6b6edd4c, node_name=llm_task)
      |  -> MapOperator(node_id=ce768e92-5299-43ab-ac04-09755b5f19e2)
      |    -> JoinOperator(node_id=f33f239e-90dd-4dd4-bc12-30d6d0912cfb)
       -> StreamingLLMOperator(node_id=ca529d57-1bd7-479c-9047-5cb5512ade8e, node_name=streaming_llm_task)
         -> OpenAIStreamingOutputOperator(node_id=1952037e-775a-4c9c-a8c3-7ca678f15871)
           -> JoinOperator(node_id=f33f239e-90dd-4dd4-bc12-30d6d0912cfb)

DAG graph

[csunny]

Great, LGTM 🚀

The DAG Visualize is cool.

[Aries-ckt]

r+