Merge pull request modelscope#3 from FredericW/main

persist function added.

examples/conversation_with_RAG_agents/agent_config.json

@@ -3,7 +3,7 @@
     "class": "LlamaIndexAgent",
     "args": {
       "name": "AgentScope Tutorial Assistant",
-      "sys_prompt": "You're a helpful assistant. You need to generate answers based on the provided context.",
+      "sys_prompt": "You're an assistant helping new users to use AgentScope. You need to generate answers based on the provided context. You only provide the most relevant information. The answer is limited to be less than 100 words. If the key words of the question can be found in the provided context, the answer should contain the section which contains the answer, for example, saying 'You may refer to SECTION_NAME for more details.' The answers must to be itemized.",
       "model_config_name": "qwen_config",
       "emb_model_config_name": "qwen_emb_config",
       "rag_config": {
@@ -22,15 +22,16 @@
             "chunk_overlap": 40,
             "similarity_top_k": 10,
             "log_retrieval": false,
-            "recent_n_mem": 1
+            "recent_n_mem": 1,
+            "persist_dir": "../../rag_storage/tutorial_assist"
       }
     }
   },
   {
     "class": "LlamaIndexAgent",
     "args": {
       "name": "AgentScope Framework Code Assistant",
-      "sys_prompt": "You're a helpful assistant about coding. You can very familiar with the framework code of AgentScope.",
+      "sys_prompt": "You're an assistant answering coding questions about AgentScope. The answer is sturucted as follows. (1) the key features, (2) an example showing how to use it, (3) where to find the example, and source codes.",
       "model_config_name": "qwen_config",
       "emb_model_config_name": "qwen_emb_config",
       "rag_config": {
@@ -63,15 +64,16 @@
             "chunk_overlap": 40,
             "similarity_top_k": 10,
             "log_retrieval": false,
-            "recent_n_mem": 1
+            "recent_n_mem": 1,
+            "persist_dir": "../../rag_storage/code_assist"
       }
     }
   },
   {
     "class": "DialogAgent",
     "args": {
       "name": "Summarize Assistant",
-      "sys_prompt": "You are a helpful assistant that can summarize the answers of the previous two messages.",
+      "sys_prompt": "You summarize the answers of the previous two messages and remove the redundant information. The answer need to be simple and itemized. The answer needs to be less than 100 words.ex .",
       "model_config_name": "qwen_config",
       "use_memory": true
     }

examples/conversation_with_RAG_agents/rag/llama_index_rag.py

@@ -6,6 +6,7 @@
 
 from typing import Any, Optional, List, Union
 from loguru import logger
+import os.path
 
 try:
     from llama_index.core.readers.base import BaseReader
@@ -21,6 +22,8 @@
     from llama_index.core.node_parser import SentenceSplitter
     from llama_index.core import (
         VectorStoreIndex,
+        StorageContext,
+        load_index_from_storage,
     )
 except ImportError:
     BaseReader, BaseRetriever = None, None
@@ -136,9 +139,8 @@ def __init__(
         super().__init__(model, emb_model, config, **kwargs)
         self.retriever = None
         self.index = None
-        self.persist_dir = kwargs.get("persist_dir", "/")
+        self.persist_dir = self.config.get("persist_dir", "/")
         self.emb_model = emb_model
-        print(self.config)
 
         # ensure the emb_model is compatible with LlamaIndex
         if isinstance(emb_model, ModelWrapperBase):
@@ -251,21 +253,39 @@ def store_and_index(
         # https://docs.llamaindex.ai/en/stable/module_guides/loading/ingestion_pipeline/root.html
         transformations.append(self.emb_model)
 
-        if vector_store is not None:
-            pipeline = IngestionPipeline(
-                transformations=transformations,
-                vector_store=vector_store,
-            )
-            _ = pipeline.run(docs)
-            self.index = VectorStoreIndex.from_vector_store(vector_store)
+        if not os.path.exists(self.persist_dir):
+            # check if index is persisted, if not, calculate the index
+            if vector_store is None:
+                # No vector_store is provide,
+                # use in memory to construct an index
+                pipeline = IngestionPipeline(
+                    transformations=transformations,
+                )
+                nodes = pipeline.run(documents=docs)
+                self.index = VectorStoreIndex(
+                    nodes=nodes,
+                    embed_model=self.emb_model,
+                )
+            else:
+                # use vector_store to construct an index
+                pipeline = IngestionPipeline(
+                    transformations=transformations,
+                    vector_store=vector_store,
+                )
+                _ = pipeline.run(docs)
+                self.index = VectorStoreIndex.from_vector_store(
+                    vector_store=vector_store,
+                )
+            # persist the calculated index
+            self.index.storage_context.persist(persist_dir=self.persist_dir)
         else:
-            # No vector store is provide, use simple in memory
-            pipeline = IngestionPipeline(
-                transformations=transformations,
-            )
-            nodes = pipeline.run(documents=docs)
-            self.index = VectorStoreIndex(
-                nodes=nodes,
+            # load the storage_context
+            storage_context = StorageContext.from_defaults(
+                persist_dir=self.persist_dir,
+                )
+            # construct index from
+            self.index = load_index_from_storage(
+                storage_context=storage_context,
                 embed_model=self.emb_model,
             )