API: Add min_tokens

Bans the EOS token until the generation reaches a minimum length. This will not prevent the model from otherwise ending the generation early by outputting other stop conditions.

sampler_overrides/sample_preset.yml

@@ -11,6 +11,9 @@
 max_tokens:
   override: 150
   force: false
+min_tokens:
+  override: 0
+  force: false
 stop:
   override: []
   force: false