Update babi_app.py

babi_app.py

@@ -15,7 +15,7 @@ from langchain.callbacks.manager import CallbackManagerForLLMRun
 from langchain.chat_models.base import BaseChatModel
 from typing import Any, Iterator, List, Optional
 from huggingface_hub import login
-# from elevenlabs import set_api_key
+
 from tempfile import TemporaryDirectory
 from langchain_community.tools.eleven_labs.text2speech import ElevenLabsText2SpeechTool
 from langchain.utilities.serpapi import SerpAPIWrapper
@@ -23,22 +23,8 @@ from langchain_community.embeddings import HuggingFaceEmbeddings
 from langchain_community.llms import HuggingFaceHub
 warnings.filterwarnings('ignore')
 
-from credits import (
-    HUGGINGFACE_TOKEN,
-    HUGGINGFACE_TOKEN as HUGGINGFACEHUB_API_TOKEN,
-    HUGGINGFACE_EMAIL,
-    HUGGINGFACE_PASS,
-    OPENAI_API_KEY,
-    ELEVENLABS_API_KEY,
-    SERPAPI_API_KEY)
-
-os.environ["HUGGINGFACEHUB_API_TOKEN"] = HUGGINGFACEHUB_API_TOKEN
-os.environ["HUGGINGFACE_TOKEN"] = HUGGINGFACE_TOKEN
-os.environ["HUGGINGFACE_EMAIL"] = HUGGINGFACE_EMAIL
-os.environ["HUGGINGFACE_PASS"] = HUGGINGFACE_PASS
-os.environ["SERPAPI_API_KEY"] = SERPAPI_API_KEY
-os.environ["OPENAI_API_KEY"] = OPENAI_API_KEY
-os.environ["ELEVEN_API_KEY"] = ELEVENLABS_API_KEY
+
+HUGGINGFACE_TOKEN = os.getenv["HUGGINGFACE_TOKEN"]
 
 if os.path.exists("/home/codemonkeyxl/.cache/huggingface/token"):
     newsession_bool = False
@@ -47,7 +33,7 @@ else:
     newsession_bool = True
     write_permission_bool = False
 
-login(HUGGINGFACEHUB_API_TOKEN, new_session= newsession_bool, write_permission= write_permission_bool )
+login(HUGGINGFACE_TOKEN, new_session= newsession_bool, write_permission= write_permission_bool )
 
 import langchain
 from langchain.chains import LLMChain
@@ -63,11 +49,9 @@ import streamlit as st
 
 from langchain.llms.base import LLM
 
-##set_api_key(ELEVENLABS_API_KEY)
 tts = ElevenLabsText2SpeechTool()
 serp_search = SerpAPIWrapper()
 
-from credits import HUGGINGFACE_TOKEN
 
 embeddings= HuggingFaceEmbeddings(
                                 model_name="all-MiniLM-L6-v2",
@@ -75,9 +59,10 @@ embeddings= HuggingFaceEmbeddings(
                                 encode_kwargs = {'normalize_embeddings': True}
                             )
 
-openllm= HuggingFaceHub( repo_id="openchat/openchat_3.5", task="text-generation", model_kwargs = {"min_length": 16,"max_length":1000,"temperature":0.1, "max_new_tokens":512, "num_return_sequences":1 })
+openllm= HuggingFaceHub( repo_id="openchat/openchat_3.5", task="text-generation", 
+                            model_kwargs = {"min_length": 16,"max_length":1000,"temperature":0.1, "max_new_tokens":512, "num_return_sequences":1 })
 best_llm = HuggingFaceHub(repo_id="tiiuae/falcon-7b-instruct", task="text-generation", 
-                          model_kwargs = {"min_length": 200,"max_length":1000,"temperature":0.1, "max_new_tokens":512, "num_return_sequences":1})           
+                              model_kwargs = {"min_length": 200,"max_length":1000,"temperature":0.1, "max_new_tokens":512, "num_return_sequences":1})           
 
 # Set Variables
 load_dotenv()
@@ -218,7 +203,6 @@ class Message:
 
 class BabyAGI(BaseModel):
     """Controller model for the BabyAGI agent."""
-
     objective: str = Field(alias="objective")
     task_list: deque = Field(default_factory=deque)
     task_creation_chain: TaskCreationChain = Field(...)
@@ -317,31 +301,31 @@ class BabyAGI(BaseModel):
 
     @classmethod
     def from_llm_and_objectives(
-        cls,
-        llm: BaseLLM,
-        vectorstore: VectorStore,
-        objective: str,
-        first_task: str,
-        verbose: bool = False,
-    ) -> "BabyAGI":
-        """Initialize the BabyAGI Controller."""
-        task_creation_chain = TaskCreationChain.from_llm(
-            llm, objective, verbose=verbose
-        )
-        task_prioritization_chain = TaskPrioritizationChain.from_llm(
-            llm, objective, verbose=verbose
-        )
-        execution_chain = ExecutionChain.from_llm(llm, vectorstore, verbose=verbose)
-        controller =  cls(
-            objective=objective,
-            task_creation_chain=task_creation_chain,
-            task_prioritization_chain=task_prioritization_chain,
-            execution_chain=execution_chain,
-        )
-        #task_id = int(time.time())
-        #controller.add_task({"task_id": task_id, "task_name": first_task})
-        controller.add_task({"task_id": 1, "task_name": first_task})
-        return controller
+            cls,
+            llm: BaseLLM,
+            vectorstore: VectorStore,
+            objective: str,
+            first_task: str,
+            verbose: bool = False,
+        ) -> "BabyAGI":
+            """Initialize the BabyAGI Controller."""
+            task_creation_chain = TaskCreationChain.from_llm(
+                llm, objective, verbose=verbose
+            )
+            task_prioritization_chain = TaskPrioritizationChain.from_llm(
+                llm, objective, verbose=verbose
+            )
+            execution_chain = ExecutionChain.from_llm(llm, vectorstore, verbose=verbose)
+            controller =  cls(
+                objective=objective,
+                task_creation_chain=task_creation_chain,
+                task_prioritization_chain=task_prioritization_chain,
+                execution_chain=execution_chain,
+            )
+            #task_id = int(time.time())
+            #controller.add_task({"task_id": task_id, "task_name": first_task})
+            controller.add_task({"task_id": 1, "task_name": first_task})
+            return controller
 
 
 def main():
@@ -361,7 +345,7 @@ def main():
     max_iterations = st.number_input("Max iterations", value=3, min_value=1, step=1)
     button = st.button("Run")
     
-    embedding_model = HuggingFaceInferenceAPIEmbeddings(api_key=os.environ["HUGGINGFACE_TOKEN"])
+    embedding_model = HuggingFaceInferenceAPIEmbeddings(api_key=os.getenv["HUGGINGFACE_TOKEN"])
 
     vectorstore = FAISS.from_texts(["_"], embedding_model, metadatas=[{"task":first_task}])