my-project-public

repository

loading code, commits, and activity

repositories

loading repo index

my-project-public — gitlawb

#1	import argparse
#2	import os
#3
#4	from src.langmem import LangMemManager
#5	from src.memzero.add import MemoryADD
#6	from src.memzero.search import MemorySearch
#7	from src.openai.predict import OpenAIPredict
#8	from src.rag import RAGManager
#9	from src.utils import METHODS, TECHNIQUES
#10	from src.zep.add import ZepAdd
#11	from src.zep.search import ZepSearch
#12
#13
#14	class Experiment:
#15	def __init__(self, technique_type, chunk_size):
#16	self.technique_type = technique_type
#17	self.chunk_size = chunk_size
#18
#19	def run(self):
#20	print(f"Running experiment with technique: {self.technique_type}, chunk size: {self.chunk_size}")
#21
#22
#23	def main():
#24	parser = argparse.ArgumentParser(description="Run memory experiments")
#25	parser.add_argument("--technique_type", choices=TECHNIQUES, default="mem0", help="Memory technique to use")
#26	parser.add_argument("--method", choices=METHODS, default="add", help="Method to use")
#27	parser.add_argument("--chunk_size", type=int, default=1000, help="Chunk size for processing")
#28	parser.add_argument("--output_folder", type=str, default="results/", help="Output path for results")
#29	parser.add_argument("--top_k", type=int, default=30, help="Number of top memories to retrieve")
#30	parser.add_argument("--filter_memories", action="store_true", default=False, help="Whether to filter memories")
#31	parser.add_argument("--is_graph", action="store_true", default=False, help="Whether to use graph-based search")
#32	parser.add_argument("--num_chunks", type=int, default=1, help="Number of chunks to process")
#33
#34	args = parser.parse_args()
#35
#36	# Add your experiment logic here
#37	print(f"Running experiments with technique: {args.technique_type}, chunk size: {args.chunk_size}")
#38
#39	if args.technique_type == "mem0":
#40	if args.method == "add":
#41	memory_manager = MemoryADD(data_path="dataset/locomo10.json", is_graph=args.is_graph)
#42	memory_manager.process_all_conversations()
#43	elif args.method == "search":
#44	output_file_path = os.path.join(
#45	args.output_folder,
#46	f"mem0_results_top_{args.top_k}_filter_{args.filter_memories}_graph_{args.is_graph}.json",
#47	)
#48	memory_searcher = MemorySearch(output_file_path, args.top_k, args.filter_memories, args.is_graph)
#49	memory_searcher.process_data_file("dataset/locomo10.json")
#50	elif args.technique_type == "rag":
#51	output_file_path = os.path.join(args.output_folder, f"rag_results_{args.chunk_size}_k{args.num_chunks}.json")
#52	rag_manager = RAGManager(data_path="dataset/locomo10_rag.json", chunk_size=args.chunk_size, k=args.num_chunks)
#53	rag_manager.process_all_conversations(output_file_path)
#54	elif args.technique_type == "langmem":
#55	output_file_path = os.path.join(args.output_folder, "langmem_results.json")
#56	langmem_manager = LangMemManager(dataset_path="dataset/locomo10_rag.json")
#57	langmem_manager.process_all_conversations(output_file_path)
#58	elif args.technique_type == "zep":
#59	if args.method == "add":
#60	zep_manager = ZepAdd(data_path="dataset/locomo10.json")
#61	zep_manager.process_all_conversations("1")
#62	elif args.method == "search":
#63	output_file_path = os.path.join(args.output_folder, "zep_search_results.json")
#64	zep_manager = ZepSearch()
#65	zep_manager.process_data_file("dataset/locomo10.json", "1", output_file_path)
#66	elif args.technique_type == "openai":
#67	output_file_path = os.path.join(args.output_folder, "openai_results.json")
#68	openai_manager = OpenAIPredict()
#69	openai_manager.process_data_file("dataset/locomo10.json", output_file_path)
#70	else:
#71	raise ValueError(f"Invalid technique type: {args.technique_type}")
#72
#73
#74	if __name__ == "__main__":
#75	main()
#76

z6Mkq5mY3JWtxoxUobWcfNHm7AkRubgSWEZTkBVqZXJviFZ5/my-project-public