ai-engineering-from-scratch/phases/17-infrastructure-and-production/28-self-hosted-serving-selection/quiz.json at main · Amjad-AbuRmileh/ai-engineering-from-scratch · 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
{
  "lesson": "28-self-hosted-serving-selection",
  "title": "Self-Hosted Serving Selection — llama.cpp, Ollama, TGI, vLLM, SGLang",
  "questions": [
    {
      "stage": "pre",
      "question": "Which engine does the lesson pick as the dev-laptop one-command default?",
      "options": [
        "TGI",
        "Ollama",
        "vLLM",
        "llama.cpp"
      ],
      "correct": 1,
      "explanation": ""
    },
    {
      "stage": "check",
      "question": "What 2025 event changes the default away from TGI for new projects?",
      "options": [
        "TGI was acquired by Anthropic",
        "TGI raised prices",
        "TGI dropped CUDA support",
        "TGI entered maintenance mode on December 11, 2025 — only bug fixes going forward"
      ],
      "correct": 3,
      "explanation": ""
    },
    {
      "stage": "check",
      "question": "Which hardware constraint forces llama.cpp and excludes vLLM / TRT-LLM?",
      "options": [
        "Apple M4",
        "AMD MI300X",
        "NVIDIA Hopper",
        "CPU only (no accelerator)"
      ],
      "correct": 3,
      "explanation": ""
    },
    {
      "stage": "check",
      "question": "Which engine does the lesson position for agentic multi-turn and prefix-heavy workloads thanks to RadixAttention?",
      "options": [
        "Ollama",
        "llama.cpp",
        "SGLang",
        "TGI"
      ],
      "correct": 2,
      "explanation": ""
    },
    {
      "stage": "post",
      "question": "What dev-to-prod pipeline does the lesson recommend on the same GGUF or HF weights?",
      "options": [
        "Only TRT-LLM, top to bottom",
        "Ollama in dev and Ollama in prod",
        "Ollama in dev, llama.cpp in staging, vLLM (or SGLang for prefix-heavy) in prod",
        "TGI everywhere"
      ],
      "correct": 2,
      "explanation": ""
    },
    {
      "stage": "post",
      "question": "Why is Ollama discouraged for shared production?",
      "options": [
        "It is closed source",
        "It only runs on Windows",
        "Go HTTP serialization adds overhead, concurrency management is simpler than vLLM, and OpenTelemetry support lags",
        "It cannot load GGUF"
      ],
      "correct": 2,
      "explanation": ""
    }
  ]
}