Context window handling in NLP - Interactive Code Practice

Practice - 5 Tasks

1fill in blank

easy

Complete the code to split text into chunks of fixed size for context window processing.

NLP

def split_text(text, chunk_size):
    return [text[i:i+[1]] for i in range(0, len(text), chunk_size)]

Drag options to blanks, or click blank then click option'

Atext

Blen(text)

Cchunk_size

Di

Attempts:

3 left

💡 Hint

Common Mistakes

Using the entire text length instead of chunk size.

Using the loop variable i as the slice size.

The chunk size determines the length of each text segment for the context window.

2fill in blank

medium

Complete the code to create overlapping context windows with a given stride.

NLP

def sliding_windows(text, window_size, stride):
    return [text[i:i+window_size] for i in range(0, len(text) - window_size + 1, [1])]

Drag options to blanks, or click blank then click option'

Awindow_size

Blen(text)

Ci

Dstride

Attempts:

3 left

💡 Hint

Common Mistakes

Using window size as the step, which causes no overlap.

Using the loop variable i as the step size.

The stride controls how much the window moves forward each step, allowing overlap.

3fill in blank

hard

Fix the error in the code that tries to pad a context window to a fixed size.

NLP

def pad_window(window, fixed_size, pad_token):
    if len(window) < fixed_size:
        window += [1] * (fixed_size - len(window))
    return window

Drag options to blanks, or click blank then click option'

Awindow

Bpad_token

Cfixed_size

Dlen(window)

Attempts:

3 left

💡 Hint

Common Mistakes

Trying to add the window to itself instead of padding.

Using fixed_size or length as the padding token.

Padding uses the pad_token repeated to fill the window to fixed size.

4fill in blank

hard

Fill both blanks to create a dictionary mapping each chunk to its start index in the text.

NLP

def chunk_indices(text, chunk_size):
    return {text[i:i+[1]]: i for i in range(0, len(text), [2])}

Drag options to blanks, or click blank then click option'

Achunk_size

Blen(text)

Cchunk_size // 2

Di

Attempts:

3 left

💡 Hint

Common Mistakes

Using different values for slice length and step size causing errors.

Using half chunk size for step causing overlapping.

Both blanks use chunk_size: slice length and step size for non-overlapping chunks.

5fill in blank

hard

Fill all three blanks to filter chunks longer than min_length and map them to their lengths.

NLP

def filter_chunks(chunks, min_length):
    return {chunk: len(chunk) for chunk in chunks if len(chunk) [1] [2] and chunk [3] ''}

Drag options to blanks, or click blank then click option'

A>

Bmin_length

C!=

D==

Attempts:

3 left

💡 Hint

Common Mistakes

Using == instead of != to check for non-empty strings.

Using < instead of > for length comparison.

Filter chunks longer than min_length and not empty strings.