AI Engineering Buildcamp: from RAG to Agents Cohort 3

Homework 1/2: RAG Statistics

Distribution of scores and reported study time for this homework.

Submissions

19

Median total score

5

Average total score

5

Score distribution

All values are points.

Questions score

Min
4
Median
5.0
Max
6
Q1
5.0
Avg
5.2
Q3
5.5

Learning in public score

Min
-
Median
0.0
Max
1
Q1
0.0
Avg
0.1
Q3
0.0

Total score

Min
4
Median
5.0
Max
6
Q1
5.0
Avg
5.3
Q3
6.0

Time distribution

All values are hours reported by students.

Lectures

Min
2.0
Median
5.0
Max
8.0
Q1
4.0
Avg
5.1
Q3
6.0

Homework

Min
1.0
Median
2.5
Max
5.0
Q1
2.0
Avg
2.7
Q3
3.4

Question breakdown

Correctness and answer distribution per question.

1. How many lines are in the extracted content from the "Think Python" book after converting to markdown?

17 / 19 correct (89.5%)

1 12,268 1 (5.3%)
2 14,268 1 (5.3%)
3 16,268 17 (89.5%)
4 18,268 0 (0.0%)

2. How many chunks are produced for the "Think Python" book with size=100 and step=50?

19 / 19 correct (100.0%)

1 134 0 (0.0%)
2 214 19 (100.0%)
3 294 0 (0.0%)
4 374 0 (0.0%)

3. How many documents (chunks) did you index in total from all books?

7 / 19 correct (36.8%)

1 719 0 (0.0%)
2 919 7 (36.8%)
3 1119 12 (63.2%)
4 1319 0 (0.0%)

4. When searching for "python function definition", which book did the top result come from?

19 / 19 correct (100.0%)

1 Think Python 19 (100.0%)
2 Think DSP 0 (0.0%)
3 Think Java 0 (0.0%)
4 Think Complexity 0 (0.0%)

5. How many input tokens did we use for the RAG query "python function definition"?

18 / 19 correct (94.7%)

1 4889 1 (5.3%)
2 6889 18 (94.7%)
3 8889 0 (0.0%)
4 10889 0 (0.0%)

6. How many MORE input tokens does the structured output version use compared to the unstructured version?

18 / 19 correct (94.7%)

1 24 0 (0.0%)
2 224 18 (94.7%)
3 424 1 (5.3%)
4 624 0 (0.0%)

Calculated: 4 May 2026, 14:57