Learn Before
Sequence Count as a Stopping Criterion in Beam Search
For the beam search algorithm, which explores multiple hypotheses in parallel, a common stopping criterion is to halt the process once a specified number of complete sequences have been found. This approach ensures that the search concludes after generating a sufficient quantity of candidate outputs.
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
End-of-Sequence (EOS) Token as a Stopping Criterion
Sequence Count as a Stopping Criterion in Beam Search
Maximum Output Length as a Stopping Criterion
Cost-Based Stopping Criteria
Behavior-Based Stopping Criteria
Debugging an Uncontrolled Text Generation System
A developer is testing a new text-generation system. They find that when prompted, the system produces a relevant initial response but then continues to generate a long, rambling stream of unrelated text until it is manually interrupted. What is the most fundamental problem with the system's configuration that leads to this behavior?
Consequences of Unbounded Text Generation
Learn After
A language model is configured to generate text using a parallel search algorithm. The generation process is set to terminate as soon as 3 complete output sequences have been found. What is the primary disadvantage of relying solely on this termination condition?
Configuring Ad Slogan Generation
Trade-offs in Generation Stopping