Update README.md

README.md

@@ -60,6 +60,8 @@ Use it to tag non-furry images.  It might not work terribly well but it might al
 
 This model may contain biases.  Tags that are poorly tagged in the original data may be weakly predicted by the classifier, for instance.  Tags that are very commonly present alongside other tags may be hallucinated.
 
+The model has been known to show biases towards English defaultness, specifically outputting the tag `english_text` on text that does not belong to any specific language, for example Arabic numerals and onomatopoeia.
+
 ### Recommendations
 
 <!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
@@ -86,6 +88,8 @@ Loss weighting was informed by a Bayesian prior model trained on a set of tags f
 
 Images go in, logits come out.  You can't explain that.
 
+Loss objective is `F.binary_cross_entropy_with_logits(output, target, torch.maximum(target, 1.0 - prior_output))`.
+
 #### Preprocessing [optional]
 
 Image preprocessing should be done in the following order:
@@ -116,19 +120,13 @@ Image preprocessing should be done in the following order:
 
 <!-- This should link to a Dataset Card if possible. -->
 
-[More Information Needed]
-
-#### Factors
-
-<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
-
-[More Information Needed]
+A validation set of approximately 128,000 images was reserved for validation testing.
 
 #### Metrics
 
 <!-- These are the evaluation metrics being used, ideally with a description of why. -->
 
-[More Information Needed]
+Bookmaker's Informedness at thresholds of 0.2, 0.3, and 0.5, as well as loss on the validation set were monitored throughout training.  Training was terminated at 5 epochs, and the checkpoint with the lowest validation loss (end of 4th epoch) was taken.
 
 ### Results