[GemmaConverter] use user_defined_symbols

src/transformers/convert_slow_tokenizer.py

@@ -1319,7 +1319,10 @@ def tokenizer(self, proto):
             raise Exception(
                 "You're trying to run a `Unigram` model but you're file was trained with a different algorithm"
             )
-
+        user_defined_symbols = [
+            AddedToken(token, normalized=False, special=False) for token in proto.trainer_spec.user_defined_symbols
+        ]
+        tokenizer.add_tokens(user_defined_symbols)
         return tokenizer
 
 

tests/models/llama/test_tokenization_llama.py

@@ -52,7 +52,7 @@
 @require_sentencepiece
 @require_tokenizers
 class LlamaTokenizationTest(TokenizerTesterMixin, unittest.TestCase):
-    from_pretrained_id = "hf-internal-testing/llama-tokenizer"
+    from_pretrained_id = ["hf-internal-testing/llama-tokenizer", "meta-llama/Llama-2-7b-hf"]
     tokenizer_class = LlamaTokenizer
     rust_tokenizer_class = LlamaTokenizerFast
 

tests/test_tokenization_common.py

@@ -51,6 +51,7 @@
     get_tests_dir,
     is_pt_tf_cross_test,
     require_jinja,
+    require_read_token,
     require_tf,
     require_tokenizers,
     require_torch,
@@ -180,6 +181,7 @@ def check_subword_sampling(
 
 class TokenizerTesterMixin:
     tokenizer_class = None
+    pretrained_tokenizer_to_test = None
     rust_tokenizer_class = None
     test_slow_tokenizer = True
     test_rust_tokenizer = True
@@ -200,13 +202,17 @@ class TokenizerTesterMixin:
     def setUp(self) -> None:
         # Tokenizer.filter makes it possible to filter which Tokenizer to case based on all the
         # information available in Tokenizer (name, rust class, python class, vocab key name)
+        self.from_pretrained_id = list(self.from_pretrained_id)
+
+        self.tokenizers_list = []
         if self.test_rust_tokenizer:
             self.tokenizers_list = [
                 (
                     self.rust_tokenizer_class,
-                    self.from_pretrained_id,
+                    pretrained_id,
                     self.from_pretrained_kwargs if self.from_pretrained_kwargs is not None else {},
                 )
+                for pretrained_id in self.from_pretrained_id
             ]
         else:
             self.tokenizers_list = []
@@ -1544,6 +1550,56 @@ def test_maximum_encoding_length_pair_input(self):
                     self.assertEqual(len(overflowing_tokens), 2 + stride)
                     self.assertEqual(overflowing_tokens, seq1_tokens[-(2 + stride) :])
 
+    @slow
+    @require_read_token
+    def test_encode_decode_fast_slow_all_tokens(self):
+        if self.rust_tokenizer_class is not None:
+            pretrained_name = self.from_pretrained_id
+
+            slow_tokenizer = self.tokenizer_class.from_pretrained(pretrained_name, legacy=False)
+            with self.subTest(f"{pretrained_name}"):
+                rust_tokenizer = self.rust_tokenizer_class.from_pretrained(
+                    pretrained_name, from_slow=True, legacy=False
+                )
+                input_full_vocab_ids = list(
+                    range(len(slow_tokenizer))
+                )  # TODO let's maybe shuffle this! And run it 4 times. This way we cover more cmbinations
+                input_full_vocab_string = rust_tokenizer.convert_tokens_to_string(
+                    rust_tokenizer.convert_ids_to_tokens(input_full_vocab_ids)
+                )
+                print(f"Length of the input string that is tested: {len(input_full_vocab_string)}")
+
+                for chunk in range(0, len(input_full_vocab_string) - 1024, 1024):
+                    string_to_check = input_full_vocab_string[chunk : chunk + 1024]
+                    with self.subTest(f"{(chunk/len(input_full_vocab_string))*100}%"):
+                        slow_encode = slow_tokenizer.encode(string_to_check)
+                        fast_encode = rust_tokenizer.encode(string_to_check)
+                        self.assertEquals(
+                            slow_encode,
+                            fast_encode,
+                            "Hint: the following tokenization diff were obtained for slow vs fast:\n "
+                            f"elements in slow: {set(slow_tokenizer.tokenize(string_to_check))-set(rust_tokenizer.tokenize(string_to_check))} \nvs\n "
+                            f"elements in fast: {set(rust_tokenizer.tokenize(string_to_check))-set(slow_tokenizer.tokenize(string_to_check))} \n"
+                            f"string used     : {string_to_check}",
+                        )
+                print(f"Length of the input ids that is tested: {len(input_full_vocab_ids)}")
+                for chunk in range(0, len(input_full_vocab_ids) - 100, 100):
+                    ids_to_decode = input_full_vocab_ids[chunk : chunk + 100]
+                    with self.subTest(f"{(chunk/len(input_full_vocab_string))*100}%"):
+                        self.assertEquals(
+                            slow_tokenizer.decode(
+                                ids_to_decode,
+                                space_between_special_tokens=False,
+                                clean_up_tokenization_spaces=False,
+                            ),
+                            rust_tokenizer.decode(
+                                ids_to_decode,
+                                space_between_special_tokens=False,
+                                clean_up_tokenization_spaces=False,
+                            ),
+                            f"Hint here are the tokens being decoded.: {slow_tokenizer.convert_ids_to_tokens(ids_to_decode)}",
+                        )
+
     # def test_encode_input_type(self):
     #     tokenizers = self.get_tokenizers(do_lower_case=False)
     #     for tokenizer in tokenizers: