Model Optimization Insights
Pradeep and Myle discuss the impact of using stale models in beam search and the effectiveness of sequence level losses on model performance. They highlight the importance of model strength in relation to the benefits of applying sequence level losses, shedding light on the challenges and advantages of large-scale experimentation in industry research.In this clip
From this podcast
NLP Highlights
53 - Classical Structured Prediction Losses for Sequence to Sequence Learning, with Sergey and Myle
Related Questions