Support min_tokens parameter

triton-inference-server · Apr 5, 2024 · 25511c3 · 25511c3
1 parent c1c88fa
commit 25511c3
Showing 1 changed file with 1 addition and 1 deletion.
diff --git a/src/model.py b/src/model.py
@@ -205,7 +205,7 @@ def get_sampling_params_dict(self, params_json):
             if k in params_dict:
                 params_dict[k] = float(params_dict[k])
 
-        int_keys = ["best_of", "max_tokens", "n", "top_k"]
+        int_keys = ["best_of", "max_tokens", "min_tokens", "n", "top_k"]
         for k in int_keys:
             if k in params_dict:
                 params_dict[k] = int(params_dict[k])