HackerLLMBench/improved_config.yaml at main · Blu1K7/HackerLLMBench · 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
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
# Improved Configuration for Benchmarking All Models
# Replace ngrok URLs with your actual endpoints

# Evaluation settings
testing_file: "evaluation/combined.jsonl"
output_dir: "scores"
log_file: "improved_testbench.log"

# Model configurations - add all your models here
models:
  # ===== LLAMA 3.1 8B =====
  # - model_id: "llama-3.1-8b-base"
  #   model_type: "llama-3.1-8b-base"
  #   endpoint: "YOUR_LLAMA_BASE_NGROK_URL/v1/chat/completions"
  #   temperature: 0.7
  #   max_tokens: 1024
  #   json_mode: true
  #   enabled: false  # Set to true when ready
  #   custom_payload: {}

  # - model_id: "llama-3.1-8b-finetuned"
  #   model_type: "llama-3.1-8b-finetuned"
  #   endpoint: "YOUR_LLAMA_FT_NGROK_URL/v1/chat/completions"
  #   temperature: 0.7
  #   max_tokens: 1024
  #   json_mode: true
  #   enabled: false
  #   custom_payload: {}

  # - model_id: "llama-3.1-8b-finetuned-600-steps"
  #   model_type: "llama-3.1-8b-finetuned-600-steps"
  #   endpoint: "YOUR_LLAMA_CKPT_NGROK_URL/v1/chat/completions"
  #   temperature: 0.7
  #   max_tokens: 1024
  #   json_mode: true
  #   enabled: false
  #   custom_payload:
  #     merged_tag: "checkpoint-600-merged"

  # - model_id: "llama-3.1-8b-finetuned-750-steps"
  #   model_type: "llama-3.1-8b-finetuned-750-steps"
  #   endpoint: "YOUR_LLAMA_CKPT_NGROK_URL/v1/chat/completions"
  #   temperature: 0.7
  #   max_tokens: 1024
  #   json_mode: true
  #   enabled: false
  #   custom_payload:
  #     merged_tag: "checkpoint-750-merged"

  # ===== GEMMA 3 12B =====
  - model_id: "gemma-3-12b-base"
    model_type: "gemma-3-12b-base"
    endpoint: "https://5eda5082011a.ngrok-free.app/v1/chat/completions"
    temperature: 1.0
    max_tokens: 1024
    json_mode: true
    enabled: false
    custom_payload:
      top_k: 64

  - model_id: "gemma-3-12b-finetuned"
    model_type: "gemma-3-12b-finetuned"
    endpoint: "https://b89b67e6799e.ngrok-free.app/v1/chat/completions"
    temperature: 1.0
    max_tokens: 1024
    json_mode: true
    enabled: true
    custom_payload:
      top_k: 64

  # - model_id: "gemma-3-12b-finetuned-800-steps"
  #   model_type: "gemma-3-12b-finetuned-800-steps"
  #   endpoint: "YOUR_GEMMA_CKPT_NGROK_URL/v1/chat/completions"
  #   temperature: 1.0
  #   max_tokens: 1024
  #   json_mode: true
  #   enabled: false
  #   custom_payload:
  #     merged_tag: "checkpoint-800-merged"
  #     top_k: 64

  # - model_id: "gemma-3-12b-finetuned-850-steps"
  #   model_type: "gemma-3-12b-finetuned-850-steps"
  #   endpoint: "YOUR_GEMMA_CKPT_NGROK_URL/v1/chat/completions"
  #   temperature: 1.0
  #   max_tokens: 1024
  #   json_mode: true
  #   enabled: false
  #   custom_payload:
  #     merged_tag: "checkpoint-850-merged"
  #     top_k: 64

  # # ===== QWEN3 14B =====
  # - model_id: "qwen3-14b-base"
  #   model_type: "qwen3-14b-base"
  #   endpoint: "https://df1fecaf3650.ngrok-free.app/v1/chat/completions"
  #   temperature: 0.7
  #   max_tokens: 1024
  #   json_mode: true
  #   enabled: true  # Currently active
  #   custom_payload:
  #     top_k: 20

  # - model_id: "qwen3-14b-finetuned"
  #   model_type: "qwen3-14b-finetuned"
  #   endpoint: "https://fc8487574bc0.ngrok-free.app/v1/chat/completions"
  #   temperature: 0.7
  #   max_tokens: 1024
  #   json_mode: true
  #   enabled: true  # Currently active
  #   custom_payload:
  #     top_k: 20

  # - model_id: "qwen3-14b-finetuned-800-steps"
  #   model_type: "qwen3-14b-finetuned-800-steps"
  #   endpoint: "https://c2446cadd3ef.ngrok-free.app/v1/chat/completions"
  #   temperature: 0.7
  #   max_tokens: 1024
  #   json_mode: true
  #   enabled: true
  #   custom_payload:
  #     merged_tag: "checkpoint-800-merged"
  #     top_k: 20

  # - model_id: "qwen3-14b-finetuned-850-steps"
  #   model_type: "qwen3-14b-finetuned-850-steps"
  #   endpoint: "https://c2446cadd3ef.ngrok-free.app/v1/chat/completions"
  #   temperature: 0.7
  #   max_tokens: 1024
  #   json_mode: true
  #   enabled: true
  #   custom_payload:
  #     merged_tag: "checkpoint-850-merged"
  #     top_k: 20

  # # ===== GPT-OSS 20B =====
  # - model_id: "gpt-oss-20b-base"
  #   model_type: "gpt-oss-20b-base"
  #   endpoint: "YOUR_GPT_OSS_BASE_NGROK_URL/v1/chat/completions"
  #   temperature: 1.0
  #   max_tokens: 1024
  #   json_mode: true
  #   enabled: false
  #   custom_payload:
  #     reasoning_effort: null  # Disable reasoning for faster responses

  # - model_id: "gpt-oss-20b-finetuned"
  #   model_type: "gpt-oss-20b-finetuned"
  #   endpoint: "YOUR_GPT_OSS_FT_NGROK_URL/v1/chat/completions"
  #   temperature: 1.0
  #   max_tokens: 1024
  #   json_mode: true
  #   enabled: false
  #   custom_payload:
  #     reasoning_effort: null

  # - model_id: "gpt-oss-20b-finetuned-800-steps"
  #   model_type: "gpt-oss-20b-finetuned-800-steps"
  #   endpoint: "YOUR_GPT_OSS_CKPT_NGROK_URL/v1/chat/completions"
  #   temperature: 1.0
  #   max_tokens: 1024
  #   json_mode: true
  #   enabled: false
  #   custom_payload:
  #     merged_tag: "checkpoint-800-merged"
  #     reasoning_effort: null

  # - model_id: "gpt-oss-20b-finetuned-850-steps"
  #   model_type: "gpt-oss-20b-finetuned-850-steps"
  #   endpoint: "YOUR_GPT_OSS_CKPT_NGROK_URL/v1/chat/completions"
  #   temperature: 1.0
  #   max_tokens: 1024
  #   json_mode: true
  #   enabled: false
  #   custom_payload:
  #     merged_tag: "checkpoint-850-merged"
  #     reasoning_effort: null