piercemaloney
/

llemma_7b

Text Generation

text-generation-inference

Inference Endpoints

Model card Files Files and versions Community

Pierce Maloney commited on Apr 10

Commit

b29c898

•

1 Parent(s): 392d92f

banning token 13

Files changed (1) hide show

handler.py +2 -1

handler.py CHANGED Viewed

@@ -23,7 +23,8 @@ class EndpointHandler():
         # Bad word: id 3070, 10456 corresponds to "(*", and we do not want to output a comment
-        bad_words_ids = [[3070], [313, 334], [10456]]
         input_ids = self.tokenizer.encode(inputs, return_tensors="pt")

         # Bad word: id 3070, 10456 corresponds to "(*", and we do not want to output a comment
+        # 13 is a newline character
+        bad_words_ids = [[3070], [313, 334], [10456], [13]]
         input_ids = self.tokenizer.encode(inputs, return_tensors="pt")