added similarity threshold to config env var

gpt_researcher/config/config.py

@@ -11,6 +11,7 @@ def __init__(self, config_file: str = None):
         self.config_file = os.path.expanduser(config_file) if config_file else os.getenv('CONFIG_FILE')
         self.retriever = os.getenv('RETRIEVER', "duckduckgo")
         self.embedding_provider = os.getenv('EMBEDDING_PROVIDER', 'openai')
+        self.fast_token_limit = int(os.getenv('SIMILARITY_THRESHOLD', 0.38))
         self.llm_provider = os.getenv('LLM_PROVIDER', "openai")
         self.ollama_base_url = os.getenv('OLLAMA_BASE_URL', None)
         self.fast_llm_model = os.getenv('FAST_LLM_MODEL', "gpt-3.5-turbo-16k")

gpt_researcher/context/compression.py

@@ -1,3 +1,4 @@
+import os
 from .retriever import SearchAPIRetriever
 from langchain.retrievers import (
     ContextualCompressionRetriever,
@@ -17,7 +18,7 @@ def __init__(self, documents, embeddings, max_results=5, **kwargs):
         self.documents = documents
         self.kwargs = kwargs
         self.embeddings = embeddings
-        self.similarity_threshold = 0.38
+        self.similarity_threshold = os.environ.get("SIMILARITY_THRESHOLD", 0.38)
 
     def __get_contextual_retriever(self):
         splitter = RecursiveCharacterTextSplitter(chunk_size=1000, chunk_overlap=100)