Learn Before
Output Stage in K-Best Selection
The output stage, labeled as step ③, is the final part of the K-best selection process. Based on a predefined 'beam width' (K), the top K candidates from the ranked list are selected as the final output. All other candidates with lower probabilities are discarded. For instance, with a beam width of K=3, the top three candidates ('cute', 'on', 'sick') are chosen, forming the final output beam.
0
1
Tags
Ch.5 Inference - Foundations of Large Language Models
Foundations of Large Language Models
Foundations of Large Language Models Course
Computing Sciences
Related
Example of K-Best Selection with a Beam Width of 3
Ranking Stage in Beam Search
Expansion Stage in K-Best Selection
A language model generates the following potential next words and their corresponding probabilities: 'house' (0.25), 'car' (0.40), 'boat' (0.15), 'plane' (0.18), and 'train' (0.02). If a selection process is used to keep only the top 3 most probable words, which set of words will be chosen?
A language model is generating the next word in a sentence. Arrange the following actions into the correct sequence for selecting the most promising candidates.
Ranking Stage in K-Best Selection
Output Stage in K-Best Selection
Impact of Selection Parameter on Text Generation
Learn After
A text generation system has produced the following list of potential next words, ranked from most to least probable: ['happy', 'glad', 'joyful', 'cheerful', 'content', 'pleased']. If the system is configured to select the final output using a 'beam width' of 3, which set of words will be chosen?
Analyzing a Text Generation Output
A language model generates a ranked list of potential next words:
['run', 'walk', 'sprint', 'jog', 'move']. The final output selected from this list is['run', 'walk']. The 'beam width' (K) used in this selection process must have been ____.