ci-model-test/config.yaml at main · christopherwoodall/ci-model-test · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
# config.yaml

evaluation:
  output:
    logs: "./logs/"
    reports: "./docs/"

  runs:
    # # Evaluation for OpenAI GPT-OSS 120B
    # openai-gpt-oss-120B:
    #   model: openrouter/openai/gpt-oss-120b
    #   json: true
    #   evals:
    #     # Spider diagram:
    #     # reading comprehension
    #     # - drop # 9536 samples
    #     # - detailbench
    #     # # # math
    #     # - math_500
    #     # # science
    #     # - openbookqa
    #     # cybersecurity
    #     # - cti_bench
    #     # - cti_bench_ate
    #     # - cti_bench_mcq
    #     # - cti_bench_rcm
    #     # - cti_bench_vsp

    # # Evaluation for OpenAI GPT-OSS 20B
    # openai-gpt-oss-20B:
    #   model: openrouter/openai/gpt-oss-20b
    #   json: true
    #   evals:
    #     # - drop
    #     # - detailbench
    #     - math_500
    #     - openbookqa
    #     - cti_bench
    #     - cti_bench_ate
    #     - cti_bench_mcq
    #     - cti_bench_rcm
    #     - cti_bench_vsp

    # # Evaluation for Qwen3 A22B (235B)
    # openrouter-qwen3-a22b-235b:
    #   model: openrouter/qwen/qwen3-235b-a22b
    #   json: true
    #   evals:
    #     # - drop
    #     # - detailbench
    #     - math_500
    #     - openbookqa
    #     - cti_bench
    #     - cti_bench_ate
    #     - cti_bench_mcq
    #     - cti_bench_rcm
    #     - cti_bench_vsp

    # Evaluation for Kimi K2 (32B)
    # openrouter-kimi-k2-32b:
    #   model: openrouter/moonshotai/kimi-k2
    #   json: true
    #   evals:
    #     # - drop
    #     # - detailbench
    #     - math_500
    #     # - openbookqa
    #     # - cti_bench
    #     # - cti_bench_ate
    #     # - cti_bench_mcq
    #     # - cti_bench_rcm
    #     # - cti_bench_vsp

    # # Evaluation for Meta LLAMA 4 Maverick (17B)
    openrouter-meta-llama4-maverick-17b:
      model: openrouter/meta-llama/llama-4-maverick
      json: true
      evals:
        # - drop
        # - detailbench
        # - math_500
        # - openbookqa
        - cti_bench
        # - cti_bench_ate
        # - cti_bench_mcq
        # - cti_bench_rcm
        # - cti_bench_vsp

    # # Evaluation for Deepseek Chat V3.1 (671B)
    openrouter-deepseek-chat-v3.1-671b:
      model: openrouter/deepseek/deepseek-chat-v3.1
      json: true
      evals:
        # - drop
        # - detailbench
        - math_500
        - openbookqa
        - cti_bench
        - cti_bench_ate
        - cti_bench_mcq
        - cti_bench_rcm
        - cti_bench_vsp