WWW
As a researcher wanting to evaluate the leave one phrase out scenario, I would like to compare greedy decoding on sentences and isolated words. For this I would need train multiple ASR models and evaluate them based on this training scenario. The goal would be to ompare greedy decoding to 1-gram and 3-gram language model for sentence and word level prompts.
WWW As a researcher wanting to evaluate the leave one phrase out scenario, I would like to compare greedy decoding on sentences and isolated words. For this I would need train multiple ASR models and evaluate them based on this training scenario. The goal would be to ompare greedy decoding to 1-gram and 3-gram language model for sentence and word level prompts.
AC Experimental code with a table of results