The relevance to assign the original model parameters when updating the means and variances of the original models. This effectively assigns a weight in frames to the original values, as if this number of frames had been seen during adaptation. The higher this number, the higher the number of examples seen in the adaptation data will have to be for the model to change significantly.
If you ran the AmTrain
task in rapid adaptation mode, set the Relevance
to 0
(zero). Any other value limits the effectiveness of the adaptation process.
Type: | Integer |
Default: | 500 |
Range: | 0–128000 |
Required: | No |
Configuration Section: | amadaptend module |
Example: | Relevance=1000
|
See Also: | Relevance (action parameter) |
|