update code: 更新断后恢复，支持vdb的更新和处理

metagpt/actions/minecraft/design_curriculumn.py

@@ -91,8 +91,9 @@ class DesignCurriculum(Action):
 
     def __init__(self, name="", context=None, llm=None):
         super().__init__(name, context, llm)
+        self.llm.model = "gpt-3.5-turbo"
 
-    async def generate_qa(self, events, qa_cache, qa_cache_questions_vectordb, human_msg, system_msg):
+    async def generate_qa(self, events, qa_cache, qa_cache_questions_vectordb, game_memory, human_msg, system_msg):
         """
         Generate qa for DesignTask's HumanMessage
         """
@@ -100,7 +101,7 @@ class DesignCurriculum(Action):
             events=events, human_msg=human_msg, system_msg=system_msg
         )
         logger.debug(f"Generate_qa_step1 result list is HERE: {questions_new}")
-        
+
         questions = []
         answers = []
         for question in questions_new:
@@ -128,6 +129,10 @@ class DesignCurriculum(Action):
             qa_cache_questions_vectordb.persist()
             questions.append(question)
             answers.append(answer)
+
+        game_memory.qa_cache = qa_cache
+        
+
         assert len(questions_new) == len(questions) == len(answers)
         logger.info(f"Curriculum Agent generate_qa Questions: {questions}")
         logger.info(f"Curriculum Agent generate_qa Answers: {answers}")
@@ -170,7 +175,7 @@ class DesignCurriculum(Action):
         # logger.info(f"Curriculum Agent generate_qa_step2 answer: {answer}")
         return answer
 
-    async def get_context_from_task(self, task, qa_cache, qa_cache_questions_vectordb):
+    async def get_context_from_task(self, task, qa_cache, qa_cache_questions_vectordb, game_memory):
         """
         Args: task
         Returns: context: "Question: {question}\n{answer}"
@@ -192,10 +197,12 @@ class DesignCurriculum(Action):
             with open(f"{CKPT_DIR}/curriculum/qa_cache.json", "w") as f:
                 json.dump(qa_cache, f)
             qa_cache_questions_vectordb.persist()
+
+        game_memory.qa_cache = qa_cache
         context = f"Question: {question}\n{answer}"
         return context
 
-    async def generate_context(self, task, qa_cache, qa_cache_questions_vectordb, max_retries=5):
+    async def generate_context(self, task, qa_cache, qa_cache_questions_vectordb, game_memory, max_retries=5):
         """
         Refer to the code in the voyager/agents/curriculum.py propose_next_ai_task() for implementation details.
         Returns: context
@@ -206,7 +213,8 @@ class DesignCurriculum(Action):
             raise RuntimeError("Max retries reached, failed to propose context.")
         try:
             context = await self.get_context_from_task(
-                task=task, qa_cache=qa_cache, qa_cache_questions_vectordb=qa_cache_questions_vectordb
+                task=task, qa_cache=qa_cache, qa_cache_questions_vectordb=qa_cache_questions_vectordb,
+                game_memory=game_memory,
             )  # Curriculum Agent Question: How to craft 4 wooden planks in Minecraft? & Curriculum Agent Answer: ...
             return context
         except Exception as e:
@@ -215,14 +223,17 @@ class DesignCurriculum(Action):
                 task=task,
                 qa_cache=qa_cache,
                 qa_cache_questions_vectordb=qa_cache_questions_vectordb,
+                game_memory=game_memory,
                 max_retries=max_retries - 1,
             )
 
-    async def run(self, task, qa_cache, qa_cache_questions_vectordb, human_msg, system_msg, *args, **kwargs):
+    async def run(self, task, qa_cache, qa_cache_questions_vectordb, game_memory, human_msg, system_msg, *args,
+                  **kwargs):
         logger.info(f"run {self.__repr__()}")
         # Generate curriculum-related questions and answers.
         # curriculum_qustion = await self.generate_qa_step1(events, human_msg, system_msg)
-        curriculum_context = await self.generate_context(task, qa_cache, qa_cache_questions_vectordb)
+        curriculum_context = await self.generate_context(task, qa_cache, qa_cache_questions_vectordb,
+                                                         game_memory=game_memory)
 
         # Return the generated questions and answers.
         return curriculum_context

metagpt/actions/minecraft/manage_skills.py

@@ -75,6 +75,12 @@ class AddNewSkills(Action):
             ids=[program_name],
             metadatas=[{"name": program_name}],
         )
+
+        skills[program_name] = {
+            "code": program_code,
+            "description": skill_desp,
+        }
+
         logger.debug(f"ADD_CHECK: There are {vectordb._collection.count()} skills in vectordb")
 
         with open(f"{CKPT_DIR}/skill/code/{dumped_program_name}.js", "w") as f:

metagpt/minecraft_team.py

@@ -60,6 +60,10 @@ class GameEnvironment(BaseModel, arbitrary_types_allowed=True):
     mf_instance: MineflayerEnv = Field(default_factory=MineflayerEnv)
     runtime_status: bool = False  # equal to action execution status: success or failed
 
+    vectordb: Chroma = Field(default_factory=Chroma)
+
+    qa_cache_questions_vectordb: Chroma = Field(default_factory=Chroma)
+
     @property
     def progress(self):
         # return len(self.completed_tasks) + 10 # Test only
@@ -84,27 +88,40 @@ class GameEnvironment(BaseModel, arbitrary_types_allowed=True):
     def core_inv_items_regex(self):
         return self.mf_instance.core_inv_items_regex
 
-    @property
-    def qa_cache_questions_vectordb(self):
-        return Chroma(
-            collection_name="qa_cache_questions_vectordb",
-            embedding_function=OpenAIEmbeddings(),
-            persist_directory=f"{CKPT_DIR}/curriculum/vectordb",
-        )
+    # @property
+    # def qa_cache_questions_vectordb(self):
+    #     return Chroma(
+    #         collection_name="qa_cache_questions_vectordb",
+    #         embedding_function=OpenAIEmbeddings(),
+    #         persist_directory=f"{CKPT_DIR}/curriculum/vectordb",
+    #     )
 
-    @property
-    def vectordb(self):
-        return Chroma(
-            collection_name="skill_vectordb",
-            embedding_function=OpenAIEmbeddings(),
-            persist_directory=f"{CKPT_DIR}/skill/vectordb",
-        )
+    # @property
+    # def vectordb(self):
+    #     return Chroma(
+    #         collection_name="skill_vectordb",
+    #         embedding_function=OpenAIEmbeddings(),
+    #         persist_directory=f"{CKPT_DIR}/skill/vectordb",
+    #     )
 
     def set_mc_port(self, mc_port):
         self.mf_instance.set_mc_port(mc_port)
         self.set_mc_resume()
 
     def set_mc_resume(self):
+        self.qa_cache_questions_vectordb = Chroma(
+            collection_name="qa_cache_questions_vectordb",
+            embedding_function=OpenAIEmbeddings(),
+            persist_directory=f"{CKPT_DIR}/curriculum/vectordb",
+        )
+
+        self.vectordb = Chroma(
+            collection_name="skill_vectordb",
+            embedding_function=OpenAIEmbeddings(),
+            persist_directory=f"{CKPT_DIR}/skill/vectordb",
+        )
+
+
         if CONFIG.resume:
             logger.info(f"Loading Action Developer from {CKPT_DIR}/action")
             with open(f"{CKPT_DIR}/action/chest_memory.json", "r") as f:
@@ -125,6 +142,7 @@ class GameEnvironment(BaseModel, arbitrary_types_allowed=True):
                 self.qa_cache = json.load(f)
 
             if self.vectordb._collection.count() == 0:
+                logger.info(self.vectordb._collection.count())
                 # Set vdvs for skills & qa_cache
                 skill_desps = [
                     skill["description"] for program_name, skill in self.skills.items()
@@ -140,36 +158,41 @@ class GameEnvironment(BaseModel, arbitrary_types_allowed=True):
                     metadatas=metadatas,
                 )
                 self.vectordb.persist()
+                
 
+            logger.info(self.qa_cache_questions_vectordb._collection.count())
             if self.qa_cache_questions_vectordb._collection.count() == 0:
+
                 questions = [question for question, answer in self.qa_cache.items()]
+
                 self.qa_cache_questions_vectordb.add_texts(texts=questions)
 
                 self.qa_cache_questions_vectordb.persist()
 
-        logger.info(
-            f"INIT_CHECK: There are {self.vectordb._collection.count()} skills in vectordb and {len(self.skills)} skills in skills.json."
-        )
-        # Check if Skill Manager's vectordb right using
-        assert self.vectordb._collection.count() >= len(self.skills), (
-            f"Skill Manager's vectordb is not synced with skills.json.\n"
-            f"There are {self.vectordb._collection.count()} skills in vectordb but {len(self.skills)} skills in skills.json.\n"
-            f"Did you set resume=False when initializing the manager?\n"
-            f"You may need to manually delete the vectordb directory for running from scratch."
-        )
+                logger.info(
+                    f"INIT_CHECK: There are {self.vectordb._collection.count()} skills in vectordb and {len(self.skills)} skills in skills.json."
+                )
+                # Check if Skill Manager's vectordb right using
+                assert self.vectordb._collection.count() >= len(self.skills), (
+                    f"Skill Manager's vectordb is not synced with skills.json.\n"
+                    f"There are {self.vectordb._collection.count()} skills in vectordb but {len(self.skills)} skills in skills.json.\n"
+                    f"Did you set resume=False when initializing the manager?\n"
+                    f"You may need to manually delete the vectordb directory for running from scratch."
+                )
+
+                logger.info(
+                    f"INIT_CHECK: There are {self.qa_cache_questions_vectordb._collection.count()} qa_cache in vectordb and {len(self.qa_cache)} questions in qa_cache.json."
+                )
+                assert self.qa_cache_questions_vectordb._collection.count() >= len(
+                    self.qa_cache
+                ), (
+                    f"Curriculum Agent's qa cache question vectordb is not synced with qa_cache.json.\n"
+                    f"There are {self.qa_cache_questions_vectordb._collection.count()} questions in vectordb "
+                    f"but {len(self.qa_cache)} questions in qa_cache.json.\n"
+                    f"Did you set resume=False when initializing the agent?\n"
+                    f"You may need to manually delete the qa cache question vectordb directory for running from scratch.\n"
+                )
 
-        logger.info(
-            f"INIT_CHECK: There are {self.qa_cache_questions_vectordb._collection.count()} qa_cache in vectordb and {len(self.qa_cache)} questions in qa_cache.json."
-        )
-        assert self.qa_cache_questions_vectordb._collection.count() >= len(
-            self.qa_cache
-        ), (
-            f"Curriculum Agent's qa cache question vectordb is not synced with qa_cache.json.\n"
-            f"There are {self.qa_cache_questions_vectordb._collection.count()} questions in vectordb "
-            f"but {len(self.qa_cache)} questions in qa_cache.json.\n"
-            f"Did you set resume=False when initializing the agent?\n"
-            f"You may need to manually delete the qa cache question vectordb directory for running from scratch.\n"
-        )
 
     def register_roles(self, roles: Iterable[Minecraft]):
         for role in roles:
@@ -209,6 +232,9 @@ class GameEnvironment(BaseModel, arbitrary_types_allowed=True):
     def update_skill_desp(self, skill_desp: str):
         self.skill_desp = skill_desp
 
+    async def update_qa_cache(self, qa_cache: dict):
+        self.qa_cache = qa_cache
+
     def update_chest_memory(self, events: Dict):
         """
         Input: events: Dict
@@ -517,3 +543,9 @@ class MinecraftPlayer(SoftwareCompany):
             # self.environment.memory.clear()
             # self._reset()
         return self.environment.history
+
+
+if __name__ == "__main__":
+    A = GameEnvironment()
+    a = A.qa_cache_questions_vectordb
+    print(a._persist_directory)

metagpt/roles/minecraft/action_developer.py

@@ -216,7 +216,7 @@ class ActionDeveloper(Base):
         self.perform_game_info_callback(new_skills_info, self.game_memory.append_skill)
     
     async def retrieve_skills(self, query, skills, *args, **kwargs):
-        retrieve_skills = await RetrieveSkills().run(query, skills)
+        retrieve_skills = await RetrieveSkills().run(query, skills, vectordb=self.game_memory.vectordb)
         logger.info(f"Render Action Agent system message with {len(retrieve_skills)} skills")
         self.perform_game_info_callback(retrieve_skills, self.game_memory.update_retrieve_skills)
         # return Message(content=f"{retrieve_skills}", instruct_content="retrieve_skills",

metagpt/roles/minecraft/curriculum_agent.py

@@ -20,11 +20,11 @@ class CurriculumDesigner(Base):
     """
 
     def __init__(
-        self,
-        name: str = "David",
-        profile: str = "Expertise in minecraft task design and curriculum development.",
-        goal: str = " Collect and integrate learner feedback to improve and refine educational content and pathways",
-        constraints: str = "Limited budget and resources for the development of educational content and technology tools.",
+            self,
+            name: str = "David",
+            profile: str = "Expertise in minecraft task design and curriculum development.",
+            goal: str = " Collect and integrate learner feedback to improve and refine educational content and pathways",
+            constraints: str = "Limited budget and resources for the development of educational content and technology tools.",
     ) -> None:
         super().__init__(name, profile, goal, constraints)
         # Initialize actions specific to the Action role
@@ -56,12 +56,12 @@ class CurriculumDesigner(Base):
         inventory = event["inventory"]
 
         if not any(
-            "dirt" in block
-            or "log" in block
-            or "grass" in block
-            or "sand" in block
-            or "snow" in block
-            for block in voxels
+                "dirt" in block
+                or "log" in block
+                or "grass" in block
+                or "sand" in block
+                or "snow" in block
+                for block in voxels
         ):
             biome = "underground"
 
@@ -92,8 +92,8 @@ class CurriculumDesigner(Base):
 
         # filter out optional inventory items if required
         if (
-            self.game_memory.progress
-            < self.game_memory.warm_up["optional_inventory_items"]
+                self.game_memory.progress
+                < self.game_memory.warm_up["optional_inventory_items"]
         ):
             inventory = {
                 k: v
@@ -121,7 +121,7 @@ class CurriculumDesigner(Base):
 
     # --------------------------------Design Task Prepare---------------------------------------
     async def render_design_task_human_message(
-        self, events, chest_observation, *args, **kwargs
+            self, events, chest_observation, *args, **kwargs
     ):
         """
         Returns: observation for curriculum
@@ -135,15 +135,19 @@ class CurriculumDesigner(Base):
             events=events, chest_observation=chest_observation
         )
         if self.game_memory.progress >= warm_up["context"]:
-        # if self.game_memory.progress >= 0: # TEST ONLY
+            # if self.game_memory.progress >= 0: # TEST ONLY
             human_msg = self.render_design_curriculum_human_message(
                 events=events, chest_observation=chest_observation
             ).content
             system_msg = [self.render_design_curriculum_system_message().content]
             questions, answers = await DesignCurriculum().generate_qa(
-                events=events, qa_cache=qa_cache, human_msg=human_msg, system_msg=system_msg
+                events=events, qa_cache=qa_cache,
+                qa_cache_questions_vectordb=self.game_memory.qa_cache_questions_vectordb,
+                game_memory=self.game_memory,
+                human_msg=human_msg, system_msg=system_msg
             )
             logger.debug(f"Generate_qa result is HERE: Ques: {questions}, Ans: {answers}")
+
             i = 1
             for question, answer in zip(questions, answers):
                 if "Answer: Unknown" in answer or "language model" in answer:
@@ -202,7 +206,7 @@ class CurriculumDesigner(Base):
         return SystemMessage(content=load_prompt("curriculum_qa_step1_ask_questions"))
 
     def render_design_curriculum_human_message(
-        self, events, chest_observation, *args, **kwargs
+            self, events, chest_observation, *args, **kwargs
     ):
         observation = self.render_curriculum_observation(
             events=events, chest_observation=chest_observation
@@ -213,7 +217,7 @@ class CurriculumDesigner(Base):
         return HumanMessage(content=content)
 
     def encapsule_design_curriculum_message(
-        self, events, chest_observation, *args, **kwargs
+            self, events, chest_observation, *args, **kwargs
     ):
         human_msg = self.render_design_curriculum_human_message(
             events=events, chest_observation=chest_observation, *args, **kwargs
@@ -309,8 +313,12 @@ class CurriculumDesigner(Base):
                 self, events=events, chest_observation=chest_observation
             )
         else:
+            logger.info(self.game_memory.qa_cache_questions_vectordb._collection.count())
+            logger.info(self.game_memory.vectordb._collection.count())
             context = await DesignCurriculum().run(
-                task, qa_cache, qa_cache_questions_vectordb, human_msg, system_msg, *args, **kwargs
+                task, qa_cache, qa_cache_questions_vectordb, game_memory=self.game_memory,
+                human_msg=human_msg,
+                system_msg=system_msg, *args, **kwargs
             )
         self.perform_game_info_callback(context, self.game_memory.update_context)
         return Message(