LLM Zoomcamp 2026
Homework 4: Evaluation Statistics
Distribution of scores and reported study time for this homework.
Score distribution
All values are points.
| Metric |
Min |
Q1 |
Median |
Average |
Q3 |
Max |
| Questions score |
1 |
5.0 |
6.0 |
5.2 |
6.0 |
6 |
| Learning in public score |
- |
0.0 |
0.0 |
0.7 |
1.0 |
7 |
| Total score |
1 |
5.0 |
6.0 |
5.9 |
6.0 |
14 |
Questions score
- Min
- 1
- Median
- 6.0
- Max
- 6
- Q1
- 5.0
- Avg
- 5.2
- Q3
- 6.0
Learning in public score
- Min
- -
- Median
- 0.0
- Max
- 7
- Q1
- 0.0
- Avg
- 0.7
- Q3
- 1.0
Total score
- Min
- 1
- Median
- 6.0
- Max
- 14
- Q1
- 5.0
- Avg
- 5.9
- Q3
- 6.0
Time distribution
All values are hours reported by students.
| Metric |
Min |
Q1 |
Median |
Average |
Q3 |
Max |
| Lectures |
0.0 |
3.0 |
4.0 |
5.7 |
6.0 |
50.0 |
| Homework |
1.0 |
2.0 |
3.0 |
3.5 |
4.0 |
24.0 |
Lectures
- Min
- 0.0
- Median
- 4.0
- Max
- 50.0
- Q1
- 3.0
- Avg
- 5.7
- Q3
- 6.0
Homework
- Min
- 1.0
- Median
- 3.0
- Max
- 24.0
- Q1
- 2.0
- Avg
- 3.5
- Q3
- 4.0
Question breakdown
Correctness and answer distribution per question.
1. When generating questions for the first 3 lesson pages, what is the average number of input tokens across these 3 calls?
354 / 374 correct
(94.7%)
1
140
8 (2.1%)
2
1400
354 (94.7%)
3
14000
10 (2.7%)
4
140000
1 (0.3%)
2. After running text_search for the first ground truth question, what is the filename of the first result?
321 / 374 correct
(85.8%)
1
01-agentic-rag/lessons/01-intro.md
47 (12.6%)
2
01-agentic-rag/lessons/03-rag.md
321 (85.8%)
3
01-agentic-rag/lessons/13-function-calling.md
5 (1.3%)
4
01-agentic-rag/lessons/10-rag-next-steps.md
0 (0.0%)
3. After running vector_search for the same question, what is the filename of the first result?
331 / 374 correct
(88.5%)
1
01-agentic-rag/lessons/01-intro.md
331 (88.5%)
2
01-agentic-rag/lessons/03-rag.md
24 (6.4%)
3
04-evaluation/lessons/11-evaluation-intro.md
12 (3.2%)
4
04-evaluation/lessons/12-rag-answers.md
4 (1.1%)
4. After evaluating text_search on the ground truth, what is the Hit Rate?
336 / 374 correct
(89.8%)
1
0.55
5 (1.3%)
2
0.66
11 (2.9%)
3
0.76
336 (89.8%)
4
0.88
18 (4.8%)
5. After evaluating vector_search on the ground truth, what is the MRR?
304 / 374 correct
(81.3%)
1
0.35
2 (0.5%)
2
0.45
17 (4.5%)
3
0.55
304 (81.3%)
4
0.65
46 (12.3%)
6. When evaluating hybrid_search for k values 1, 50, 100, and 200, which k gives the best MRR?
300 / 374 correct
(80.2%)
1
1
300 (80.2%)
2
50
58 (15.5%)
3
100
9 (2.4%)
4
200
1 (0.3%)
Calculated:
22 July 2026, 19:37