OpenCompass/configs/datasets/ceval/README.md

C-Eval

Base Models

model	ceval-test	ceval-test-hard	ceval-test-stem	ceval-test-social-science	ceval-test-humanities	ceval-test-other	ceval-dev	ceval-dev-hard	ceval-dev-stem	ceval-dev-social-science	ceval-dev-humanities	ceval-dev-other
llama-7b-turbomind	26.61	27.75	27.20	26.31	25.90	26.52	27.44	27.68	27.16	29.49	24.18	29.36
llama-13b-turbomind	29.18	25.59	27.66	33.86	28.29	28.58	31.75	30.32	31.39	35.22	30.16	30.82
llama-30b-turbomind	35.09	31.68	34.56	39.89	33.02	33.76	37.70	31.97	34.80	42.72	41.19	34.93
llama-65b-turbomind	37.98	29.47	36.03	45.03	36.51	36.56	40.46	33.76	36.37	46.47	42.26	40.63
llama-2-7b-turbomind	30.13	26.26	29.29	33.02	31.02	28.15	32.70	25.85	28.75	39.75	37.04	29.13
llama-2-13b-turbomind	37.38	30.81	35.85	43.98	36.81	34.75	40.43	31.34	35.67	45.75	45.32	39.36
llama-2-70b-turbomind	49.53	33.48	44.73	60.19	50.93	47.17	50.26	32.53	44.83	59.44	54.45	47.58
llama-3-8b-turbomind	48.83	34.47	46.02	56.48	49.15	46.69	50.45	33.76	45.94	58.08	50.93	51.25
llama-3-70b-turbomind	66.56	54.09	64.08	76.43	64.38	64.25	67.30	52.35	62.67	77.89	69.76	63.65
internlm2-1.8b-turbomind	44.79	33.93	41.19	54.26	47.15	40.35	46.64	33.00	38.62	57.28	51.30	46.89
internlm2-7b-turbomind	63.54	45.32	58.10	76.40	66.94	58.32	64.23	40.09	54.37	76.88	70.11	64.77
internlm2-20b-turbomind	67.28	50.15	62.33	79.59	70.55	61.82	66.73	42.50	59.25	79.98	73.43	61.56
qwen-1.8b-turbomind	54.24	38.60	50.02	68.18	55.33	48.13	53.78	33.38	46.36	68.40	57.57	50.17
qwen-7b-turbomind	62.06	42.73	56.21	77.12	65.28	55.76	63.23	36.99	54.74	78.55	68.94	59.02
qwen-14b-turbomind	70.33	53.61	65.25	83.19	72.85	65.37	72.05	55.03	66.07	85.59	74.91	67.78
qwen-72b-turbomind	83.25	66.78	78.44	91.75	83.86	83.63	83.60	63.68	78.05	90.25	87.13	84.13
qwen1.5-0.5b-hf	48.36	35.55	44.72	62.00	48.51	42.41	50.43	37.00	46.28	62.64	48.11	49.18
qwen1.5-1.8b-hf	58.67	40.98	53.91	74.52	58.51	53.06	59.38	43.02	53.45	75.88	60.06	54.47
qwen1.5-4b-hf	66.55	48.50	61.45	81.12	67.90	61.22	66.46	43.12	56.76	82.89	67.61	68.03
qwen1.5-7b-hf	72.49	52.90	66.77	85.50	74.37	69.19	73.57	49.16	66.32	84.23	77.30	73.34
qwen1.5-14b-hf	76.93	60.50	72.08	88.81	77.95	73.94	77.86	54.81	71.55	86.79	82.86	76.23
qwen1.5-32b-hf	82.50	66.67	77.97	90.93	83.66	81.88	82.79	71.06	80.01	89.02	83.36	81.62
qwen1.5-72b-hf	83.03	65.09	77.90	91.47	83.85	83.86	83.72	64.09	77.26	91.87	87.64	84.14
qwen1.5-moe-a2-7b-hf	76.67	51.37	68.89	88.33	77.15	79.73	77.90	51.25	67.27	89.28	83.16	81.60
mistral-7b-v0.1-hf	43.76	33.85	42.23	49.97	41.10	43.54	47.54	33.97	44.74	54.80	51.52	42.06
mistral-7b-v0.2-hf	42.81	32.84	41.00	50.19	39.45	42.77	46.44	31.67	42.89	54.50	48.75	43.23
mixtral-8x7b-v0.1-hf	51.15	41.46	50.93	59.19	46.69	48.72	55.31	42.04	52.78	62.00	56.44	52.71
mixtral-8x22b-v0.1-hf	58.13	48.31	58.01	66.94	53.60	54.86	60.50	45.67	57.44	71.27	61.31	55.47
yi-6b-hf	70.78	43.72	60.54	83.29	75.39	73.40	73.13	46.87	63.14	85.52	78.70	74.45
yi-34b-hf	80.93	58.51	73.48	89.24	83.65	84.18	81.62	56.95	71.64	89.73	87.49	86.53
deepseek-7b-base-hf	43.68	28.90	37.03	53.55	50.14	40.34	45.07	31.94	38.81	56.68	47.10	43.85
deepseek-67b-base-hf	66.66	44.25	57.89	79.02	72.36	65.66	66.65	38.62	56.65	79.56	73.72	66.01

Details on Test Split

model	computer_network	operating_system	computer_architecture	college_programming	college_physics	college_chemistry	advanced_mathematics	probability_and_statistics	discrete_mathematics	electrical_engineer	metrology_engineer	high_school_mathematics
llama-7b-turbomind	29.82	25.70	26.94	30.99	32.95	23.66	26.01	22.89	27.45	30.09	26.48	33.13
llama-13b-turbomind	33.33	37.99	31.09	29.82	22.16	27.23	31.79	27.11	24.84	28.02	33.33	30.72
llama-30b-turbomind	40.94	48.60	40.41	34.21	32.95	35.71	36.42	32.53	27.45	31.56	36.07	30.12
llama-65b-turbomind	41.52	50.84	44.04	40.94	27.84	29.46	28.32	30.72	29.41	35.10	42.47	30.12
llama-2-7b-turbomind	33.92	37.99	34.72	30.99	26.70	21.88	31.79	25.30	24.18	31.56	39.73	30.12
llama-2-13b-turbomind	40.94	46.93	37.82	36.26	30.68	29.46	35.84	30.72	24.84	32.74	42.92	34.94
llama-2-70b-turbomind	55.56	58.66	53.89	47.95	34.09	33.48	32.95	27.11	34.64	37.76	57.99	29.52
llama-3-8b-turbomind	55.56	58.66	55.96	51.17	27.27	35.27	36.42	31.33	34.64	40.12	50.68	30.72
llama-3-70b-turbomind	69.59	75.98	69.95	71.64	49.43	58.04	52.02	53.01	58.82	45.72	68.95	40.96
internlm2-1.8b-turbomind	40.35	40.78	39.38	32.16	34.66	34.38	31.21	31.33	35.95	35.10	51.60	27.71
internlm2-7b-turbomind	56.14	57.54	62.69	49.42	43.75	48.21	34.68	32.53	33.33	41.00	60.27	40.36
internlm2-20b-turbomind	62.57	65.36	66.84	58.77	43.18	51.79	39.31	40.36	35.95	42.77	66.67	47.59
qwen-1.8b-turbomind	46.20	41.90	46.63	36.84	40.34	36.61	27.75	28.92	32.68	36.58	57.08	30.12
qwen-7b-turbomind	52.63	54.75	54.40	46.20	35.80	44.20	36.99	27.71	26.80	38.35	57.99	33.13
qwen-14b-turbomind	58.48	64.80	59.07	54.68	45.45	57.59	45.09	33.73	39.22	49.26	67.58	45.78
qwen-72b-turbomind	83.04	73.74	79.27	76.61	75.00	64.29	49.13	44.58	46.41	66.37	85.84	68.07
qwen1.5-0.5b-hf	37.43	40.22	41.45	35.09	40.91	34.82	30.06	27.11	26.80	29.79	54.34	31.93
qwen1.5-1.8b-hf	47.37	50.84	47.67	38.30	43.18	35.27	29.48	30.12	33.99	39.53	58.90	28.92
qwen1.5-4b-hf	62.57	56.98	56.99	46.78	48.30	45.98	40.46	34.34	31.37	46.61	62.10	43.37
qwen1.5-7b-hf	66.08	62.57	66.32	55.56	54.55	47.77	41.62	31.93	35.95	49.85	74.43	49.40
qwen1.5-14b-hf	71.35	66.48	68.39	64.91	57.95	65.62	41.62	40.36	47.71	56.64	79.45	56.63
qwen1.5-32b-hf	84.80	73.18	74.61	70.18	71.59	61.61	49.13	45.78	49.02	61.95	87.67	72.89
qwen1.5-72b-hf	85.38	73.74	78.24	78.36	72.73	63.39	43.35	40.96	49.02	65.78	85.84	66.27
qwen1.5-moe-a2-7b-hf	77.78	73.74	68.91	64.91	66.48	49.11	33.53	36.75	35.95	61.06	91.32	40.96
mistral-7b-v0.1-hf	55.56	55.31	56.99	48.25	39.77	39.29	33.53	25.90	31.37	35.99	45.21	27.11
mistral-7b-v0.2-hf	56.14	53.63	55.44	47.66	36.36	34.38	32.37	25.30	33.33	31.86	45.21	29.52
mixtral-8x7b-v0.1-hf	62.57	64.80	60.10	60.53	38.64	42.41	40.46	37.35	45.75	35.99	60.27	34.94
mixtral-8x22b-v0.1-hf	65.50	74.86	63.73	65.79	46.59	52.68	52.02	45.78	52.94	42.77	62.56	39.16
yi-6b-hf	68.42	63.13	69.43	57.89	42.05	48.66	31.79	33.13	28.76	49.85	74.89	37.35
yi-34b-hf	83.63	80.45	74.09	68.42	62.50	60.27	45.09	38.55	50.33	65.19	88.58	49.40
deepseek-7b-base-hf	44.44	44.13	44.56	36.26	30.68	29.02	32.37	24.70	26.14	35.99	48.86	28.31
deepseek-67b-base-hf	63.16	70.39	65.80	59.36	42.61	45.54	35.84	38.55	42.48	44.54	68.95	33.73

model	high_school_physics	high_school_chemistry	high_school_biology	middle_school_mathematics	middle_school_biology	middle_school_physics	middle_school_chemistry	veterinary_medicine	college_economics	business_administration	marxism	mao_zedong_thought
llama-7b-turbomind	29.14	26.74	24.57	29.94	22.92	23.60	20.00	30.95	29.98	24.58	25.70	25.11
llama-13b-turbomind	22.29	18.60	28.00	26.55	26.56	25.28	19.46	29.05	28.77	28.57	39.66	43.38
llama-30b-turbomind	25.14	33.14	36.00	31.07	39.06	28.09	33.51	38.10	35.21	35.88	48.04	33.33
llama-65b-turbomind	33.71	26.16	38.29	33.90	44.27	36.52	38.92	38.10	37.42	42.19	59.22	48.40
llama-2-7b-turbomind	26.86	23.26	26.86	28.81	28.12	29.78	22.70	30.48	31.79	30.56	33.52	36.07
llama-2-13b-turbomind	28.00	31.98	36.57	36.72	38.54	36.52	37.84	46.67	37.02	36.54	57.54	41.10
llama-2-70b-turbomind	40.00	36.05	48.00	36.72	66.67	55.06	55.68	52.86	51.91	48.50	68.16	60.73
llama-3-8b-turbomind	41.71	38.37	50.86	36.16	61.98	63.48	63.78	56.19	41.65	49.17	69.27	54.34
llama-3-70b-turbomind	63.43	56.98	69.14	59.32	84.90	75.28	78.92	79.52	68.81	59.80	86.59	79.91
internlm2-1.8b-turbomind	30.29	45.93	46.29	33.33	63.02	60.11	62.70	47.62	35.61	37.87	69.27	61.64
internlm2-7b-turbomind	64.57	65.12	76.00	54.80	91.15	85.96	90.27	74.29	57.34	50.50	86.59	83.56
internlm2-20b-turbomind	68.57	74.42	78.86	58.76	91.67	90.45	90.27	72.38	57.95	55.81	88.83	88.58
qwen-1.8b-turbomind	55.43	56.98	61.14	54.80	85.42	84.83	85.41	54.76	43.06	44.19	83.80	79.91
qwen-7b-turbomind	68.00	69.19	82.86	57.63	93.75	87.64	92.43	63.81	47.28	57.48	86.59	82.65
qwen-14b-turbomind	78.86	83.14	92.57	67.23	96.88	95.51	96.76	73.33	56.94	64.45	91.62	86.76
qwen-72b-turbomind	93.14	93.60	95.43	88.70	98.44	97.75	99.46	90.00	75.45	80.73	96.09	99.54
qwen1.5-0.5b-hf	48.57	44.19	60.00	40.68	73.44	69.66	78.92	49.05	34.41	40.20	79.89	74.43
qwen1.5-1.8b-hf	58.86	68.02	76.00	59.32	91.15	90.45	87.03	63.81	44.87	48.50	86.03	90.41
qwen1.5-4b-hf	66.86	77.33	82.86	68.93	95.31	92.70	97.30	71.90	51.31	61.13	91.62	94.52
qwen1.5-7b-hf	79.43	82.56	91.43	77.40	96.88	95.51	96.22	80.00	62.37	69.77	93.30	97.26
qwen1.5-14b-hf	86.29	87.79	93.14	83.05	97.92	95.51	97.84	82.86	63.78	77.08	95.53	96.35
qwen1.5-32b-hf	88.00	95.35	94.86	91.53	97.92	99.44	100.00	90.00	73.44	78.74	94.97	98.63
qwen1.5-72b-hf	91.43	93.60	95.43	88.70	97.92	98.31	99.46	90.00	74.25	80.40	94.41	98.63
qwen1.5-moe-a2-7b-hf	70.86	77.33	82.86	68.36	97.92	93.26	97.30	89.52	70.22	74.75	96.09	98.17
mistral-7b-v0.1-hf	33.14	40.70	40.57	40.11	47.92	49.44	50.81	47.62	44.87	37.87	58.10	48.40
mistral-7b-v0.2-hf	34.86	36.63	45.71	36.72	46.35	46.07	48.65	43.81	43.46	39.53	57.54	48.86
mixtral-8x7b-v0.1-hf	49.71	42.44	53.71	47.46	62.50	61.24	60.00	57.62	52.52	44.52	68.72	57.99
mixtral-8x22b-v0.1-hf	54.29	43.02	58.29	55.93	76.04	66.29	75.68	66.19	60.97	51.83	74.30	70.78
yi-6b-hf	58.86	69.19	78.29	43.50	92.19	89.33	90.27	83.81	59.56	70.10	93.85	97.72
yi-34b-hf	80.00	81.98	93.14	65.54	97.40	95.51	96.76	92.86	74.04	76.08	94.97	97.26
deepseek-7b-base-hf	29.14	30.81	33.14	24.29	53.12	45.51	48.65	50.48	38.23	44.19	62.01	65.30
deepseek-67b-base-hf	60.00	55.23	64.00	46.33	84.90	79.78	83.24	73.33	57.75	63.79	89.94	88.58

model	education_science	teacher_qualification	high_school_politics	high_school_geography	middle_school_politics	middle_school_geography	modern_chinese_history	ideological_and_moral_cultivation	logic	law	chinese_language_and_literature	art_studies
llama-7b-turbomind	22.96	31.58	25.57	29.78	22.80	25.00	21.70	21.51	25.00	26.24	22.49	25.84
llama-13b-turbomind	29.26	30.83	33.52	36.52	34.72	33.33	24.06	40.12	26.47	33.48	30.14	29.87
llama-30b-turbomind	37.41	46.37	32.95	38.20	50.78	40.74	28.77	45.93	33.33	32.13	39.23	22.82
llama-65b-turbomind	39.63	51.13	31.82	39.89	58.03	42.59	34.91	55.23	39.71	30.32	37.80	32.89
llama-2-7b-turbomind	27.78	34.34	31.82	34.83	35.23	34.26	28.77	38.95	32.35	33.94	27.27	30.87
llama-2-13b-turbomind	41.48	47.37	37.50	37.64	50.78	52.78	43.40	48.84	32.35	38.46	36.36	30.20
llama-2-70b-turbomind	57.78	69.17	50.57	58.43	69.95	66.67	50.94	72.09	50.98	42.53	44.98	52.01
llama-3-8b-turbomind	56.30	65.41	47.16	56.18	64.25	61.11	55.66	67.44	41.67	40.27	45.45	50.34
llama-3-70b-turbomind	72.22	85.46	75.00	74.72	84.97	76.85	75.00	76.16	59.31	52.94	62.68	68.46
internlm2-1.8b-turbomind	47.41	61.40	55.11	47.75	61.66	64.81	61.79	63.95	32.35	32.58	48.33	36.58
internlm2-7b-turbomind	66.67	85.96	78.98	74.72	91.71	87.96	80.66	80.23	42.16	50.23	64.11	70.13
internlm2-20b-turbomind	69.26	89.22	83.52	80.34	90.67	91.67	83.02	85.47	49.02	54.30	72.25	73.15
qwen-1.8b-turbomind	51.11	70.68	71.02	62.36	88.60	87.04	69.81	73.26	29.90	46.15	50.24	47.32
qwen-7b-turbomind	57.41	83.71	88.64	79.78	93.26	94.44	75.47	79.07	42.16	47.96	59.33	65.10
qwen-14b-turbomind	72.96	89.97	93.75	83.71	96.37	95.37	86.32	87.21	50.00	60.63	66.99	72.48
qwen-72b-turbomind	85.56	96.24	95.45	93.26	97.93	97.22	92.45	91.86	67.65	76.92	75.12	83.89
qwen1.5-0.5b-hf	43.33	63.16	65.91	56.18	82.90	79.63	68.87	70.35	28.43	37.56	39.23	32.21
qwen1.5-1.8b-hf	57.41	76.44	81.25	75.84	92.75	91.67	79.72	81.98	34.31	47.96	47.85	43.62
qwen1.5-4b-hf	65.93	87.47	86.93	82.58	94.30	95.37	84.91	84.30	40.20	62.90	58.85	58.72
qwen1.5-7b-hf	69.26	91.98	90.91	89.89	95.85	94.44	89.15	87.21	48.04	67.87	63.16	68.12
qwen1.5-14b-hf	78.89	94.99	94.89	91.57	96.89	98.15	91.04	88.37	57.84	69.68	66.99	73.83
qwen1.5-32b-hf	83.70	95.99	93.75	94.38	98.45	97.22	90.57	91.28	70.10	76.92	76.56	80.87
qwen1.5-72b-hf	84.44	96.49	96.59	93.82	98.45	97.22	92.92	91.28	66.67	76.92	74.16	85.23
qwen1.5-moe-a2-7b-hf	80.74	95.49	89.20	89.33	94.82	94.44	92.45	91.28	52.45	75.57	67.94	79.87
mistral-7b-v0.1-hf	45.19	59.15	43.75	49.44	56.48	56.48	45.28	58.14	37.75	38.91	40.67	34.56
mistral-7b-v0.2-hf	45.93	58.65	38.07	48.31	63.21	58.33	41.98	54.07	35.78	40.27	38.28	32.21
mixtral-8x7b-v0.1-hf	57.04	67.92	53.41	55.06	69.95	64.81	47.64	70.93	42.16	38.01	46.41	36.58
mixtral-8x22b-v0.1-hf	60.37	72.68	64.77	65.17	77.20	71.30	57.08	75.00	49.51	43.44	52.63	49.33
yi-6b-hf	79.26	92.48	77.27	76.40	92.75	93.52	89.15	90.12	60.78	74.66	61.24	74.16
yi-34b-hf	84.81	96.24	88.07	88.20	96.37	96.30	91.98	91.28	75.00	78.73	80.38	82.89
deepseek-7b-base-hf	52.22	70.18	47.16	51.12	60.62	44.44	58.49	66.86	31.86	37.56	53.11	61.07
deepseek-67b-base-hf	76.67	89.22	77.27	78.65	89.64	78.70	85.85	84.30	50.00	64.25	69.38	84.23

model	professional_tour_guide	legal_professional	high_school_chinese	high_school_history	middle_school_history	civil_servant	sports_science	plant_protection	basic_medicine	clinical_medicine	urban_and_rural_planner	accountant
llama-7b-turbomind	29.70	23.72	27.53	30.22	30.92	27.04	22.78	28.64	28.00	25.00	26.32	29.80
llama-13b-turbomind	25.94	20.93	25.84	29.67	24.64	29.60	26.67	29.15	33.71	25.50	28.47	28.44
llama-30b-turbomind	29.32	27.91	30.34	36.26	37.20	36.13	36.11	38.69	34.29	29.50	38.52	29.35
llama-65b-turbomind	28.95	30.70	30.90	44.51	35.75	36.60	45.56	39.20	37.71	30.00	39.47	37.02
llama-2-7b-turbomind	29.70	30.23	24.72	29.67	34.78	30.07	31.11	31.16	30.29	25.50	31.34	27.31
llama-2-13b-turbomind	30.83	32.56	24.16	42.31	45.41	32.87	36.67	45.23	38.29	33.50	35.17	34.31
llama-2-70b-turbomind	53.76	38.14	30.34	58.79	65.70	43.82	51.11	58.29	49.71	42.00	49.76	46.28
llama-3-8b-turbomind	52.63	42.33	27.53	51.65	65.70	44.52	54.44	51.26	46.86	43.00	46.41	45.15
llama-3-70b-turbomind	72.93	52.56	32.58	71.98	83.57	56.88	69.44	78.89	76.00	67.50	57.89	59.14
internlm2-1.8b-turbomind	51.50	38.14	25.84	56.04	71.50	47.32	35.00	43.72	42.29	39.00	41.15	36.57
internlm2-7b-turbomind	72.56	53.49	52.25	79.67	90.82	62.00	62.78	64.32	66.86	59.50	55.74	53.50
internlm2-20b-turbomind	74.06	54.42	56.18	81.87	92.27	61.77	68.33	69.85	68.00	63.50	60.77	58.92
qwen-1.8b-turbomind	54.14	43.72	39.89	69.23	85.02	49.88	45.56	48.74	48.57	51.50	46.89	45.82
qwen-7b-turbomind	71.05	48.37	53.93	81.87	93.72	59.67	54.44	62.31	58.29	57.50	50.24	56.66
qwen-14b-turbomind	79.70	53.02	63.48	87.36	94.20	71.33	63.33	71.36	73.14	68.00	59.09	67.95
qwen-72b-turbomind	90.23	77.21	79.21	91.76	96.14	77.86	86.11	85.43	91.43	90.50	76.08	86.68
qwen1.5-0.5b-hf	44.36	36.74	39.33	58.24	78.26	43.36	40.00	45.23	41.71	42.50	43.54	43.12
qwen1.5-1.8b-hf	59.40	47.91	37.08	72.53	91.30	53.61	53.33	51.26	49.71	58.00	51.20	56.21
qwen1.5-4b-hf	65.04	58.60	55.62	83.52	94.20	62.00	63.89	65.33	65.71	64.00	55.26	61.40
qwen1.5-7b-hf	78.57	66.51	66.85	87.91	94.69	68.07	65.00	64.82	77.14	77.50	60.77	74.49
qwen1.5-14b-hf	83.08	72.09	70.22	90.11	94.20	69.46	73.89	70.35	82.29	83.00	65.31	78.33
qwen1.5-32b-hf	87.59	78.14	79.78	92.86	95.65	78.32	80.56	79.90	90.29	89.00	77.27	86.68
qwen1.5-72b-hf	91.35	76.74	79.21	91.76	96.14	79.25	85.56	86.93	92.00	90.00	75.84	86.91
qwen1.5-moe-a2-7b-hf	88.35	75.81	51.12	79.12	94.69	67.37	80.56	73.37	87.43	84.00	78.23	82.39
mistral-7b-v0.1-hf	40.23	39.07	24.16	41.21	52.17	41.49	45.00	52.26	45.14	42.00	42.58	44.02
mistral-7b-v0.2-hf	36.84	34.88	23.03	43.96	52.66	40.79	50.00	50.75	45.14	40.50	42.58	40.86
mixtral-8x7b-v0.1-hf	47.74	40.00	28.09	57.14	58.94	44.29	58.33	53.77	48.57	46.00	51.20	46.50
mixtral-8x22b-v0.1-hf	59.02	41.86	29.78	60.99	71.01	50.82	57.78	67.34	62.29	52.00	53.35	55.98
yi-6b-hf	85.34	67.91	53.93	80.22	91.79	65.97	72.22	72.36	82.29	84.50	69.86	71.56
yi-34b-hf	94.36	76.74	65.73	87.91	95.17	79.25	85.56	90.95	90.86	92.00	76.79	82.39
deepseek-7b-base-hf	65.79	29.30	32.58	47.80	67.15	37.76	44.44	52.26	43.43	36.50	41.15	37.02
deepseek-67b-base-hf	83.83	58.60	45.51	79.67	90.34	62.47	70.56	70.85	81.14	71.50	61.72	60.05

model	fire_engineer	environmental_impact_assessment_engineer	tax_accountant	physician
llama-7b-turbomind	22.34	24.91	29.12	27.77
llama-13b-turbomind	24.11	30.25	27.77	30.70
llama-30b-turbomind	28.72	31.67	31.83	36.57
llama-65b-turbomind	28.37	39.15	33.63	35.44
llama-2-7b-turbomind	22.70	24.91	25.51	29.80
llama-2-13b-turbomind	25.53	35.94	29.35	35.44
llama-2-70b-turbomind	36.52	52.67	36.12	52.60
llama-3-8b-turbomind	35.46	49.82	41.31	55.30
llama-3-70b-turbomind	48.58	64.41	52.60	75.40
internlm2-1.8b-turbomind	32.27	42.35	39.05	45.15
internlm2-7b-turbomind	46.81	55.16	47.63	67.27
internlm2-20b-turbomind	45.04	62.63	51.47	69.75
qwen-1.8b-turbomind	41.84	47.69	45.60	57.34
qwen-7b-turbomind	41.84	54.80	48.08	69.53
qwen-14b-turbomind	45.74	64.77	56.43	77.88
qwen-72b-turbomind	80.50	74.73	81.04	89.62
qwen1.5-0.5b-hf	39.36	41.28	38.37	48.08
qwen1.5-1.8b-hf	45.74	49.47	51.69	63.43
qwen1.5-4b-hf	50.35	51.60	58.69	75.17
qwen1.5-7b-hf	58.51	65.84	67.04	81.94
qwen1.5-14b-hf	63.83	67.26	72.23	87.36
qwen1.5-32b-hf	74.47	73.31	80.14	90.74
qwen1.5-72b-hf	79.79	75.09	81.04	90.07
qwen1.5-moe-a2-7b-hf	74.82	77.58	79.68	91.65
mistral-7b-v0.1-hf	32.27	45.91	37.70	50.56
mistral-7b-v0.2-hf	32.62	44.13	36.79	46.28
mixtral-8x7b-v0.1-hf	35.11	53.02	46.73	52.37
mixtral-8x22b-v0.1-hf	38.65	56.23	49.21	59.82
yi-6b-hf	67.38	68.68	69.53	83.07
yi-34b-hf	77.66	83.27	77.43	89.84
deepseek-7b-base-hf	30.50	38.79	35.67	46.28
deepseek-67b-base-hf	46.81	65.12	54.40	77.65

Details on Dev Split

Chat Models

model	ceval-test	ceval-test-hard	ceval-test-stem	ceval-test-social-science	ceval-test-humanities	ceval-test-other	ceval-dev	ceval-dev-hard	ceval-dev-stem	ceval-dev-social-science	ceval-dev-humanities	ceval-dev-other
qwen1.5-0.5b-chat-hf	36.88	28.83	34.49	43.46	37.35	34.76	38.58	33.90	33.63	43.81	41.79	39.59
qwen1.5-1.8b-chat-hf	55.17	38.21	50.63	70.26	56.04	48.82	55.93	37.60	50.31	67.59	60.90	50.59
qwen1.5-4b-chat-hf	61.54	44.79	56.86	75.84	62.13	56.46	62.76	38.32	55.39	79.53	65.67	58.00
qwen1.5-7b-chat-hf	68.71	51.77	64.27	81.23	68.22	65.88	71.10	50.13	65.42	83.99	73.77	67.02
qwen1.5-14b-chat-hf	74.80	56.54	69.46	87.47	76.46	71.32	76.35	52.08	69.68	86.70	80.56	74.87
qwen1.5-32b-chat-hf	80.47	63.17	75.66	89.58	81.98	79.43	81.27	63.51	76.64	89.39	82.97	80.59
qwen1.5-72b-chat-hf	81.53	63.62	75.86	90.74	83.18	81.84	82.88	62.44	77.54	89.80	86.11	83.07
qwen1.5-110b-chat-hf	87.33	67.27	80.70	93.58	89.67	91.35	87.59	73.64	81.94	91.47	92.12	89.80
internlm2-chat-1.8b-hf	47.04	34.81	43.28	59.34	48.24	41.50	48.51	36.75	42.23	57.79	54.83	45.15
internlm2-chat-1.8b-sft-hf	47.19	35.34	43.49	59.56	48.30	41.58	48.75	35.83	42.04	59.80	54.84	44.83
internlm2-chat-7b-hf	58.75	39.61	52.38	71.46	61.57	55.96	61.04	36.56	51.81	74.01	69.13	57.92
internlm2-chat-7b-sft-hf	58.96	40.09	52.40	71.49	62.20	56.26	61.02	37.29	52.60	74.01	68.27	57.27
internlm2-chat-20b-hf	63.12	42.65	56.21	75.64	67.15	60.27	63.45	34.96	52.84	79.27	71.50	60.32
internlm2-chat-20b-sft-hf	63.16	42.70	56.19	75.74	67.20	60.37	63.54	34.96	52.57	80.33	71.42	60.34
llama-3-8b-instruct-hf	50.90	34.54	46.73	58.73	49.24	53.04	52.55	36.37	48.47	58.03	53.26	54.26
llama-3-70b-instruct-hf	67.38	54.02	65.16	76.83	62.29	67.92	67.92	54.50	66.85	76.80	65.98	63.72
llama-3-8b-instruct-lmdeploy	49.92	34.75	46.19	58.49	47.68	51.14	50.27	33.32	46.25	56.93	49.02	52.76
llama-3-70b-instruct-lmdeploy	66.41	52.76	64.72	75.31	61.36	66.44	68.21	52.28	65.86	75.06	68.37	66.09
mistral-7b-instruct-v0.1-hf	36.76	27.76	35.55	42.41	34.45	36.12	40.04	30.21	35.77	45.15	40.99	42.22
mistral-7b-instruct-v0.2-hf	40.38	30.26	38.82	47.66	37.08	39.91	43.00	25.97	38.60	47.44	48.15	41.82
mixtral-8x7b-instruct-v0.1-hf	49.61	37.78	47.86	58.56	46.40	47.85	51.68	37.41	49.14	59.79	52.97	47.65

Details on Test Split

model	computer_network	operating_system	computer_architecture	college_programming	college_physics	college_chemistry	advanced_mathematics	probability_and_statistics	discrete_mathematics	electrical_engineer	metrology_engineer	high_school_mathematics
qwen1.5-0.5b-chat-hf	35.67	36.87	33.68	33.92	35.23	28.12	27.17	26.51	24.84	28.91	40.18	25.90
qwen1.5-1.8b-chat-hf	46.78	47.49	50.78	39.18	41.48	31.25	32.95	27.71	28.10	34.81	55.71	27.11
qwen1.5-4b-chat-hf	54.39	54.75	54.92	44.74	46.02	43.30	39.31	31.33	28.10	45.13	58.90	43.98
qwen1.5-7b-chat-hf	60.82	60.34	63.21	55.85	48.86	45.09	46.24	36.14	39.22	47.49	70.32	45.78
qwen1.5-14b-chat-hf	69.59	62.57	64.77	64.91	55.68	57.14	49.13	32.53	43.14	55.16	76.71	46.99
qwen1.5-32b-chat-hf	81.87	74.30	73.58	71.35	63.07	60.71	50.87	46.99	47.06	59.29	83.11	60.84
qwen1.5-72b-chat-hf	77.78	75.42	76.17	73.39	63.64	62.50	45.09	45.78	48.37	59.00	81.74	60.84
qwen1.5-110b-chat-hf	83.63	86.03	81.87	77.49	76.70	67.86	49.13	47.59	55.56	79.94	95.89	62.05
internlm2-chat-1.8b-hf	42.11	43.58	44.56	35.38	32.95	34.82	32.95	28.92	32.68	34.22	53.42	31.93
internlm2-chat-1.8b-sft-hf	42.11	44.13	43.01	35.09	34.09	36.16	32.95	27.11	33.33	35.10	51.14	33.13
internlm2-chat-7b-hf	59.65	60.89	58.03	51.46	36.93	43.75	36.99	29.52	36.60	39.82	63.47	38.55
internlm2-chat-7b-sft-hf	59.06	61.45	56.48	52.63	39.77	41.52	36.99	27.71	39.22	40.12	62.10	40.36
internlm2-chat-20b-hf	61.99	70.39	63.73	54.97	33.52	47.77	43.93	40.96	44.44	44.25	61.64	34.34
internlm2-chat-20b-sft-hf	61.40	70.39	63.21	54.97	32.95	47.77	42.20	42.17	43.14	44.25	61.64	32.53
llama-3-8b-instruct-hf	57.31	58.10	57.51	51.17	28.41	35.27	39.31	32.53	35.29	38.05	55.25	27.11
llama-3-70b-instruct-hf	71.93	74.86	70.98	67.54	50.57	57.14	52.60	53.01	56.21	47.79	68.95	43.98
llama-3-8b-instruct-lmdeploy	55.56	57.54	55.44	48.25	30.11	33.04	35.84	31.33	33.33	38.94	53.88	31.93
llama-3-70b-instruct-lmdeploy	70.76	77.09	69.95	67.84	49.43	54.02	50.87	54.22	56.21	47.20	69.86	42.17
mistral-7b-instruct-v0.1-hf	49.12	47.49	43.52	39.18	32.39	28.57	29.48	24.10	28.10	37.46	44.29	23.49
mistral-7b-instruct-v0.2-hf	47.95	53.07	52.85	42.69	28.41	26.79	40.46	30.12	29.41	33.33	42.92	24.10
mixtral-8x7b-instruct-v0.1-hf	58.48	62.57	58.03	56.43	38.64	36.16	39.31	34.94	37.91	34.81	55.71	28.31

model	high_school_physics	high_school_chemistry	high_school_biology	middle_school_mathematics	middle_school_biology	middle_school_physics	middle_school_chemistry	veterinary_medicine	college_economics	business_administration	marxism	mao_zedong_thought
qwen1.5-0.5b-chat-hf	30.86	31.98	44.00	27.68	47.40	40.45	55.14	35.24	32.80	30.56	58.66	57.53
qwen1.5-1.8b-chat-hf	54.86	62.21	69.14	53.67	82.81	83.15	85.41	58.10	44.06	49.83	82.12	82.65
qwen1.5-4b-chat-hf	58.86	67.44	80.00	55.93	89.58	88.20	88.11	64.29	47.08	57.48	86.59	84.93
qwen1.5-7b-chat-hf	72.00	80.81	84.00	70.06	95.31	94.94	95.14	73.81	56.94	66.11	91.62	89.04
qwen1.5-14b-chat-hf	84.00	83.72	90.29	80.23	97.92	94.94	98.38	81.43	63.18	74.75	93.30	96.80
qwen1.5-32b-chat-hf	85.71	90.12	93.71	85.31	97.92	98.31	100.00	89.05	69.82	75.75	93.85	97.72
qwen1.5-72b-chat-hf	88.57	94.19	94.86	85.31	97.92	97.75	98.38	90.48	71.63	79.73	93.85	97.72
qwen1.5-110b-chat-hf	86.86	92.44	94.29	85.31	98.44	98.88	98.92	95.24	78.87	86.38	95.53	99.54
internlm2-chat-1.8b-hf	35.43	48.84	52.00	35.03	70.31	67.98	67.03	41.43	37.83	36.88	70.95	60.73
internlm2-chat-1.8b-sft-hf	37.71	48.26	53.14	34.46	71.35	67.98	67.57	41.90	38.63	37.54	72.63	60.27
internlm2-chat-7b-hf	46.29	48.26	60.57	46.89	78.65	71.91	71.35	68.10	50.30	50.83	77.09	76.26
internlm2-chat-7b-sft-hf	46.86	48.26	61.14	45.76	77.60	71.91	71.35	67.62	50.10	50.50	77.09	75.80
internlm2-chat-20b-hf	49.71	46.51	63.43	55.37	80.73	74.72	79.46	72.38	55.73	59.80	85.47	76.26
internlm2-chat-20b-sft-hf	53.71	47.09	64.00	55.37	80.73	73.60	78.92	73.81	55.53	60.13	85.47	75.80
llama-3-8b-instruct-hf	38.86	39.53	50.29	40.11	65.10	60.11	63.78	61.43	47.89	45.85	69.27	56.16
llama-3-70b-instruct-hf	63.43	55.23	69.71	68.36	85.42	80.90	78.38	86.19	69.01	65.12	83.24	82.65
llama-3-8b-instruct-lmdeploy	41.71	40.70	52.00	41.24	61.46	58.43	65.41	57.62	45.27	46.18	69.27	55.71
llama-3-70b-instruct-lmdeploy	61.71	53.49	70.86	64.97	88.02	83.71	77.30	84.76	68.21	60.80	80.45	79.91
mistral-7b-instruct-v0.1-hf	27.43	28.49	36.00	28.25	40.10	42.70	43.78	37.14	32.80	37.87	41.90	48.86
mistral-7b-instruct-v0.2-hf	33.14	29.65	44.00	31.07	47.92	44.94	49.19	44.29	37.02	40.86	53.63	48.40
mixtral-8x7b-instruct-v0.1-hf	46.29	40.70	54.86	42.37	58.85	60.67	57.84	54.29	50.10	46.51	69.27	52.51

model	education_science	teacher_qualification	high_school_politics	high_school_geography	middle_school_politics	middle_school_geography	modern_chinese_history	ideological_and_moral_cultivation	logic	law	chinese_language_and_literature	art_studies
qwen1.5-0.5b-chat-hf	33.33	46.12	37.50	37.08	57.51	43.52	42.45	51.74	32.84	31.22	37.32	24.50
qwen1.5-1.8b-chat-hf	54.07	72.43	74.43	66.85	89.12	87.04	77.36	76.16	38.24	44.34	46.89	40.94
qwen1.5-4b-chat-hf	60.00	84.71	82.39	69.66	94.82	90.74	79.72	78.49	41.67	57.47	54.07	56.38
qwen1.5-7b-chat-hf	66.30	90.73	84.66	80.90	94.30	91.67	82.55	84.88	38.73	60.18	60.77	63.42
qwen1.5-14b-chat-hf	74.81	93.73	90.91	92.13	96.89	98.15	89.62	88.37	54.41	70.14	69.86	69.13
qwen1.5-32b-chat-hf	80.37	94.49	93.75	94.94	97.93	97.22	90.09	90.70	68.63	78.73	73.21	77.52
qwen1.5-72b-chat-hf	84.07	96.74	95.45	94.94	97.93	95.37	92.92	91.28	63.73	80.09	73.68	83.89
qwen1.5-110b-chat-hf	90.37	96.99	96.02	95.51	98.45	98.15	93.87	94.19	81.37	86.88	84.69	90.94
internlm2-chat-1.8b-hf	48.15	65.41	69.32	54.49	79.27	70.37	60.85	64.53	32.35	32.58	45.45	40.60
internlm2-chat-1.8b-sft-hf	48.15	64.91	69.89	53.93	79.27	70.37	61.32	63.95	33.82	29.86	45.45	39.93
internlm2-chat-7b-hf	66.67	85.21	73.30	66.85	91.19	76.85	70.28	75.58	42.16	50.68	60.77	70.47
internlm2-chat-7b-sft-hf	67.04	85.21	73.86	66.85	90.67	77.78	71.70	75.00	42.16	51.13	60.29	72.15
internlm2-chat-20b-hf	74.07	85.96	75.57	77.53	89.12	76.85	72.64	83.72	51.96	56.11	68.42	73.49
internlm2-chat-20b-sft-hf	73.70	85.46	76.70	78.09	89.64	76.85	72.17	84.88	50.00	56.56	66.99	75.17
llama-3-8b-instruct-hf	55.93	67.42	55.68	55.06	72.02	62.04	54.25	66.86	44.12	40.72	47.37	44.63
llama-3-70b-instruct-hf	71.11	84.21	74.43	73.03	84.97	80.56	69.81	78.49	57.35	50.68	57.89	64.43
llama-3-8b-instruct-lmdeploy	54.81	67.17	58.52	53.37	72.54	62.04	57.08	63.95	44.12	37.56	46.89	42.62
llama-3-70b-instruct-lmdeploy	70.37	82.96	72.16	71.91	83.94	82.41	69.34	77.91	55.39	50.68	56.46	64.09
mistral-7b-instruct-v0.1-hf	39.63	46.62	33.52	41.01	56.48	45.37	36.32	43.60	29.90	31.67	39.71	31.88
mistral-7b-instruct-v0.2-hf	46.30	54.39	39.20	43.26	61.66	51.85	35.38	55.23	28.92	35.29	37.80	29.19
mixtral-8x7b-instruct-v0.1-hf	58.52	66.17	56.82	57.30	66.32	62.04	48.11	66.28	41.67	37.10	46.41	35.91

model	professional_tour_guide	legal_professional	high_school_chinese	high_school_history	middle_school_history	civil_servant	sports_science	plant_protection	basic_medicine	clinical_medicine	urban_and_rural_planner	accountant
qwen1.5-0.5b-chat-hf	36.47	39.07	27.53	41.76	45.89	39.63	35.56	31.66	37.71	34.00	32.78	37.25
qwen1.5-1.8b-chat-hf	56.02	45.58	39.33	67.03	84.54	49.42	48.89	51.76	47.43	50.50	45.69	52.14
qwen1.5-4b-chat-hf	61.28	52.56	42.70	73.08	85.99	55.48	59.44	55.28	60.57	57.00	50.00	58.01
qwen1.5-7b-chat-hf	73.31	56.28	58.99	82.97	88.41	64.57	66.67	63.82	77.14	75.50	57.42	69.07
qwen1.5-14b-chat-hf	80.83	65.12	70.79	89.56	93.24	67.60	72.78	68.34	80.57	80.00	61.72	75.62
qwen1.5-32b-chat-hf	87.59	72.56	76.40	90.66	95.65	74.36	80.00	80.40	86.86	84.00	74.88	85.33
qwen1.5-72b-chat-hf	90.98	76.28	75.84	90.66	95.65	75.52	84.44	82.91	91.43	89.00	73.92	85.10
qwen1.5-110b-chat-hf	95.11	88.37	82.58	91.76	96.62	87.65	91.67	90.95	93.71	95.00	87.08	91.87
internlm2-chat-1.8b-hf	54.14	40.00	27.53	62.09	70.53	44.99	41.67	51.76	45.71	39.00	40.67	39.28
internlm2-chat-1.8b-sft-hf	54.14	42.33	26.97	61.54	71.98	45.45	41.67	50.25	45.14	37.50	41.39	40.63
internlm2-chat-7b-hf	70.68	44.19	34.83	73.63	84.06	51.98	57.22	68.34	66.86	57.50	54.55	50.11
internlm2-chat-7b-sft-hf	71.80	44.65	37.64	73.63	84.06	51.98	57.78	67.84	65.71	60.50	54.55	50.11
internlm2-chat-20b-hf	75.56	54.42	42.13	74.73	85.51	57.34	65.56	67.84	73.71	64.00	57.89	55.98
internlm2-chat-20b-sft-hf	76.32	55.35	41.01	75.27	85.51	58.28	65.56	67.34	72.57	65.00	58.37	56.43
llama-3-8b-instruct-hf	53.01	44.65	33.15	46.70	66.18	45.22	58.89	61.81	62.86	57.50	48.33	49.89
llama-3-70b-instruct-hf	71.43	50.70	30.90	71.43	82.13	59.67	73.33	73.37	82.86	82.00	59.09	62.08
llama-3-8b-instruct-lmdeploy	51.13	45.12	29.78	43.96	62.32	47.09	56.11	54.77	56.00	56.00	49.04	47.40
llama-3-70b-instruct-lmdeploy	68.80	48.84	30.90	70.88	81.64	58.28	72.22	70.85	80.00	81.00	57.66	62.53
mistral-7b-instruct-v0.1-hf	30.45	35.81	24.72	40.11	34.78	30.77	43.89	38.69	36.57	32.50	44.74	34.09
mistral-7b-instruct-v0.2-hf	36.09	38.14	23.03	43.41	45.41	35.90	50.00	41.71	42.86	36.00	45.22	42.21
mixtral-8x7b-instruct-v0.1-hf	47.37	44.65	30.34	51.65	60.87	42.19	53.89	58.29	52.00	47.00	48.56	44.02

model	fire_engineer	environmental_impact_assessment_engineer	tax_accountant	physician
qwen1.5-0.5b-chat-hf	27.66	38.43	32.28	35.44
qwen1.5-1.8b-chat-hf	38.65	46.62	46.73	59.14
qwen1.5-4b-chat-hf	49.29	54.80	51.02	70.20
qwen1.5-7b-chat-hf	53.90	62.28	57.79	76.52
qwen1.5-14b-chat-hf	58.87	65.12	67.27	86.68
qwen1.5-32b-chat-hf	74.11	70.82	74.94	88.04
qwen1.5-72b-chat-hf	74.82	75.09	78.56	89.39
qwen1.5-110b-chat-hf	88.30	88.97	94.13	95.49
internlm2-chat-1.8b-hf	30.14	41.99	34.54	46.73
internlm2-chat-1.8b-sft-hf	30.14	43.06	34.31	47.86
internlm2-chat-7b-hf	42.20	52.31	47.63	66.82
internlm2-chat-7b-sft-hf	43.26	52.67	47.86	66.59
internlm2-chat-20b-hf	45.74	54.80	51.02	69.07
internlm2-chat-20b-sft-hf	45.74	55.16	51.02	68.62
llama-3-8b-instruct-hf	37.59	50.53	42.44	68.40
llama-3-70b-instruct-hf	50.71	64.06	55.53	84.42
llama-3-8b-instruct-lmdeploy	37.94	50.53	41.53	66.14
llama-3-70b-instruct-lmdeploy	48.94	63.70	53.95	81.72
mistral-7b-instruct-v0.1-hf	27.66	39.15	29.35	39.95
mistral-7b-instruct-v0.2-hf	32.27	37.01	32.96	42.89
mixtral-8x7b-instruct-v0.1-hf	36.88	48.75	41.76	53.05

Details on Dev Split