דסבת

Discuss, Learn and be Happy דיון בשאלות

• «
• 1
• 2
• ...
• 23
• 24
• 25
• 26
• 27
• 28
• 29
• ...
• 36
• 37
• »

How is Matrix Rank defined?

-

1

+

שאלה #134489

		The number of columns in the matrix
		The total number of elements in the matrix
done		The number of linearly independent columns (or rows) in the matrix
		The size of the vector space generated by the matrix columns

נערך לפני 5 חודשים Shachar Adam

• gavel
• mood_bad
• share

הגב

מיין לפי דירוג חדש ישן

What is Low Rank Matrix decomposition?

-

1

+

שאלה #134490

		Decomposing a matrix into its row and column vectors
		Decomposing a matrix into its singular values
done		Representing a matrix as a multiplication of two smaller matrices, one with linearly independent rows and the other with linearly independent columns
		Representing a matrix as a sum of smaller matrices with orthogonal rows and columns

• gavel
• mood_bad
• share

הגב

מיין לפי דירוג חדש ישן

What defines Linearly Dependent Vectors in a vector space?

-

1

+

שאלה #134491

		A set of vectors with equal magnitudes but different directions
		A set of vectors that are all perpendicular to the same plane
		A set of vectors where each vector is orthogonal to the others
done		A set of vectors where at least one vector can be expressed as a linear combination of others or obtained through scalar multiplication

נערך לפני 5 חודשים Shachar Adam

• gavel
• mood_bad
• share

הגב

מיין לפי דירוג חדש ישן

How does LoRA leverage the low intrinsic dimensionality of pretrained models during fine-tuning?

-

1

+

שאלה #134492

		By compressing the pretrained model into a smaller size before fine-tuning
		By discarding the original model parameters and initializing new ones from scratch
done		By representing the change in weights as a product of two low-rank matrices and updating the original weights with these matrices

• gavel
• mood_bad
• share

הגב

מיין לפי דירוג חדש ישן

What is one of the main advantages of LoRA?

-

1

+

שאלה #134508

done		Significant reduction in computational complexity
		Slightly lower memory consumption

• gavel
• mood_bad
• share

הגב

מיין לפי דירוג חדש ישן

What is the purpose of creating a constant length dataset for training?

-

1

+

שאלה #134519

		To increase the computational complexity of the training process
		To introduce variability in the length of training examples
done		To simplify training by ensuring examples have a fixed context size
		To reduce the efficiency of the training algorithm

• gavel
• mood_bad
• share

הגב

מיין לפי דירוג חדש ישן

Why is padding and truncation considered wasteful in traditional data preparation methods?

-

1

+

שאלה #134520

		Because it increases the computational efficiency during the forward pass
		Because it prevents bias in the model by removing unnecessary tokens
		Because it ensures all training examples have identical lengths
done		Because it may lead to loss of data and computational inefficiency

• gavel
• mood_bad
• share

הגב

מיין לפי דירוג חדש ישן

What is the alternative approach to padding and truncation in data preparation?

-

1

+

שאלה #134521

sentiment_very_satisfied		Tokenizing the data and concatenating with an end-of-sequence token
		Reducing the number of training examples
		Adding more layers to the model architecture
		Removing the end-of-sequence token from training examples

נערך לפני 5 חודשים Shachar Adam

• gavel
• mood_bad
• share

הגב

מיין לפי דירוג חדש ישן

What is Natural Language Processing (NLP)?

-

1

+

שאלה #135496

done		subfield of artificial intelligence and linguistics that studies problems related to the processing and manipulation of natural language.
		A technique for translating text from one language to another.
		A method for analyzing computer programming languages.
		A discipline focused on studying the behavior of animals in their natural habitats.

• gavel
• mood_bad
• share

הגב

מיין לפי דירוג חדש ישן

What is a “quintillion”?

-

1

+

שאלה #135500

		A 1 followed by 9 zeros
		A 1 followed by 12 zeros
		A 1 followed by 15 zeros
done		A 1 followed by 18 zeros

• gavel
• mood_bad
• share

הגב

מיין לפי דירוג חדש ישן

• «
• 1
• 2
• ...
• 23
• 24
• 25
• 26
• 27
• 28
• 29
• ...
• 36
• 37
• »