Reputation: 735
I am new to Sphinx 4. I have been reading about Sphinx 4 and have to implement it as a part of the project. I am building upon HelloNGram model. To improve accuracy I wish to include a language model I downloaded from http://www.keithv.com/software/giga/ I have downloaded 64k NVP 3 gram.
I see these following files in the zip archive lm_giga_64k_nvp.hdecode.dic lm_giga_64k_nvp.hvite.dic lm_giga_64k_nvp.sphinx.dic lm_giga_64k_nvp.sphinx.filler lm_giga_64k_nvp_3gram.arpa phonelist wlist_giga_64k_nvp wlist_giga_64k_nvp_start_end
I want to integrate it in HelloNGram. How do I go about doing this? Please explain all the steps. I am very new to sphinx 4! Or if there is any other model you would suggest and can instruct as to how to integrate it in my project... Thanks!
Upvotes: 1
Views: 1429
Reputation: 25220
Step 1. Convert lm_giga_64k_nvp_3gram.arpa to DMP format with sphinx_lm_convert from sphinxbase.
Step 2. In config file change the type of language model component from SimpleNGramModel to LargeTrigramModel
Step 3. Update component property to load converted DMP file
To learn more about Sphinx4 read the tutorial
http://cmusphinx.sourceforge.net/wiki/tutorial
It will help you to answer the questions you have
Upvotes: 3