Update README.md

README.md

@@ -57,10 +57,16 @@ With streaming, the results with different chunk sizes on the test split are the
 |       | full   | cs=32 (1280ms) | 24 (960ms) | 16 (640ms) | 12 (480ms) | 8 (320ms) |
 |:-----:|:------:|:------:|:------:|:------:|:------:|:------:|
 | full  | 11.00% | -      | -      | -      | -      | -      |
-| 16    | -      | -      | -      | TBD    | 11.84% | TBD    |
-| 8     | -      | -      | 11.50% | TBD    | TBD    | TBD    |
-| 4     | -      | 11.40% | 11.53% | TBD    | TBD    | TBD    |
-| 2     | -      | 11.46% | 11.67% | TBD    | TBD    | TBD    |
+| 16    | -      | -      | -      | 11.70% | 11.84% | 12.14% |
+| 8     | -      | -      | 11.50% | 11.72% | 11.88% | 12.28% |
+| 4     | -      | 11.40% | 11.53% | 11.81% | 12.03% | 12.64% |
+| 2     | -      | 11.46% | 11.67% | 12.03% | 12.43% | 13.25% |
+| 1\*   | -      | 11.59% | TBD    | TBD    | TBD    | TBD    |
+
+(\*: model was never explicitly trained with this setting)
+
+When comparing configurations, bear in mind that the utterances of the test set are of a finite length.  
+All WER values were evaluated using [this script](https://gist.github.com/asumagic/66aaf43156c31866cf2ebcffda46ea01) (until we integrate a nicer way to evaluate streaming WER in different conditions).
 
 ## Pipeline description