Moonlight-16B-A3B-Instruct Benchmark Details

Moonlight-16B-A3B-Instruct currently shows benchmark results led by MBPP (23 / 70, score 63.80), GSM8K (26 / 70, score 77.40), HumanEval (54 / 101, score 48.10).

Benchmark Results

Moonlight-16B-A3B-Instruct

Benchmark Results

General Knowledge

3 evaluations

Benchmark / mode

Score

Rank/total

MMLU

Standard Mode

73 / 124

BBH

Standard Mode

65.20

15 / 21

MMLU Pro

Standard Mode

42.40

128 / 133

Math and Reasoning

2 evaluations

Benchmark / mode

Score

Rank/total

GSM8K

Standard Mode

77.40

26 / 70

MATH

Standard Mode

45.30

36 / 42

Coding and Software Engineer

2 evaluations

Benchmark / mode

Score

Rank/total

MBPP

Standard Mode

63.80

23 / 70

HumanEval

Standard Mode

48.10

54 / 101

Compare with other models