An LLM-powered knowledge curation system that researches a topic and generates a full-length report with citations.
Python agentic-ragdeep-researchemnlp2024knowledge-curationlarge-language-modelsnaaclnlpreport-generationretrieval-augmented-generation
Update run_costorm_gpt.py
Yijia Shao4c16d2e9e194
diff --git a/examples/costorm_examples/run_costorm_gpt.py b/examples/costorm_examples/run_costorm_gpt.pyindex 8fe6c7f..66d6bd3 100644--- a/examples/costorm_examples/run_costorm_gpt.py+++ b/examples/costorm_examples/run_costorm_gpt.py@@ -1,5 +1,5 @@"""-STORM Wiki pipeline powered by GPT-4o/4o-mini and Bing search engine.+Co-STORM pipeline powered by GPT-4o/4o-mini and Bing search engine.You need to set up the following environment variables to run this script:- OPENAI_API_KEY: OpenAI API key- OPENAI_API_TYPE: OpenAI API type (e.g., 'openai' or 'azure')@@ -116,6 +116,9 @@ def main(args):# warm start the systemcostorm_runner.warm_start()+ # Below is an example of how users may interact with Co-STORM to seek information together+ # In actual deployment, we suggest allowing the user to decide whether to observe the agent utterance or inject a turn+# observing Co-STORM LLM agent utterance for 5 turnsfor _ in range(1):conv_turn = costorm_runner.step()@@ -235,4 +238,4 @@ if __name__ == '__main__':help='If set, enable console log print.')- main(parser.parse_args())\ No newline at end of file+ main(parser.parse_args())