results/LLaMA-2-7B-32k.json

{"passkey": {"4k": 1.0, "8k": 1.0, "16k": 1.0, "32k": 1.0, "64k": 0, "128k": 0, "dataset_total_score": 0.6666666666666666}, "matreshka_yes_no": {"4k": 0.5016722408026756, "8k": 0.5, "16k": 0.5, "32k": 0.5, "64k": 0, "128k": 0, "dataset_total_score": 0.33361204013377926}, "matreshka_names": {"4k": 0.08, "8k": 0.06666666666666667, "16k": 0.02, "32k": 0.04, "64k": 0, "128k": 0, "dataset_total_score": 0.034444444444444444}, "passkey_with_librusec": {"4k": 1.0, "8k": 0.975, "16k": 0.985, "32k": 0.97, "64k": 0, "128k": 0, "dataset_total_score": 0.6549999999999999}, "librusec_history": {"8k": 0.6875, "16k": 0.5, "32k": 0.4375, "64k": 0, "dataset_total_score": 0.40625}, "ru_gsm100": {"16k": 0.07, "dataset_total_score": 0.07}, "ru_sci_passage_count": {"4k": 0.18, "8k": 0.05, "16k": 0.05, "32k": 0.005, "64k": 0, "128k": 0, "dataset_total_score": 0.047499999999999994}, "ru_2wikimultihopqa": {"8k": 0.4489795918367347, "16k": 0.3984375, "32k": 0.2682926829268293, "dataset_total_score": 0.37190325825452136}, "long_context_multiq": {"8k": 0.33, "16k": 0.1, "4k": 0.045, "32k": 0.0, "64k": 0, "128k": 0, "dataset_total_score": 0.07916666666666666}, "ru_sci_abstract_retrieval": {"4k": 0.8519047619047619, "8k": 0.7612582259010829, "16k": 0.4675868475624726, "32k": 0.2665782908189807, "64k": 0, "128k": 0, "dataset_total_score": 0.39122135436454974}, "ru_trec": {"4k": 0.24324324324324326, "8k": 0.18, "16k": 0.24175824175824176, "32k": 0.28688524590163933, "dataset_total_score": 0.2379716827257811}, "ru_sci_fi": {"32k": 0.1111111111111111, "64k": 0, "dataset_total_score": 0.05555555555555555}, "librusec_mhqa": {"8k": 0.2760416666666667, "dataset_total_score": 0.2760416666666667}, "ru_babilong_qa1": {"4k": 0.6, "8k": 0.66, "16k": 0.66, "32k": 0.5, "64k": 0, "128k": 0, "dataset_total_score": 0.4033333333333333}, "ru_babilong_qa2": {"4k": 0.25, "8k": 0.3, "16k": 0.25875, "32k": 0.19, "64k": 0, "128k": 0, "dataset_total_score": 0.16645833333333335}, "ru_babilong_qa3": {"4k": 0.22933333333333333, "8k": 0.28933333333333333, "16k": 0.26, "32k": 0.2, "64k": 0, "128k": 0, "dataset_total_score": 0.1631111111111111}, "ru_babilong_qa4": {"4k": 0.31, "8k": 0.34, "16k": 0.23, "32k": 0.12, "64k": 0, "128k": 0, "dataset_total_score": 0.16666666666666666}, "ru_babilong_qa5": {"4k": 0.59, "8k": 0.66, "16k": 0.64, "32k": 0.69, "64k": 0, "128k": 0, "dataset_total_score": 0.43}, "ru_quality": {"16k": 0.13871635610766048, "8k": 0.17073170731707318, "dataset_total_score": 0.15472403171236682}, "ru_tpo": {"8k": 0.5431606905710492, "dataset_total_score": 0.5431606905710492}, "ru_qasper": {"16k": 0.05999038960490889, "8k": 0.0580459343880765, "32k": 0.022401950361811175, "dataset_total_score": 0.04681275811826552}, "total_score": 0.2714095362059408}