Fix moses punct (#140)

* Fix moses punctuation

* max_number actually not there - but it's in the generation_config!

Author: johnml1135

machine/translation/huggingface/hugging_face_nmt_engine.py

@@ -55,8 +55,8 @@ def __init__(
         self._tokenizer = AutoTokenizer.from_pretrained(self._model.name_or_path, use_fast=True)
         if isinstance(self._tokenizer, (NllbTokenizer, NllbTokenizerFast)):
             self._mpn = MosesPunctNormalizer()
-            self._mpn.substitutions = [
-                (str(re.compile(r)), sub)
+            self._mpn.substitutions = [  # type: ignore
+                (re.compile(r), sub)
                 for r, sub in self._mpn.substitutions
                 if isinstance(r, str) and isinstance(sub, str)
             ]
@@ -236,8 +236,12 @@ def _forward(self, model_inputs, **generate_kwargs):
 
         input_tokens = model_inputs["input_tokens"]
         del model_inputs["input_tokens"]
-        generate_kwargs["min_length"] = generate_kwargs.get("min_length", self.model.config.min_length)
-        generate_kwargs["max_length"] = generate_kwargs.get("max_length", self.model.config.max_length)
+        if hasattr(self.model, "generation_config") and self.model.generation_config is not None:
+            config = self.model.generation_config
+        else:
+            config = self.model.config
+        generate_kwargs["min_length"] = generate_kwargs.get("min_length", config.min_length)
+        generate_kwargs["max_length"] = generate_kwargs.get("max_length", config.max_length)
         self.check_inputs(input_length, generate_kwargs["min_length"], generate_kwargs["max_length"])
         output = self.model.generate(
             **model_inputs,

machine/translation/huggingface/hugging_face_nmt_model_trainer.py

@@ -100,10 +100,8 @@ def __init__(
         self._add_unk_src_tokens = add_unk_src_tokens
         self._add_unk_tgt_tokens = add_unk_tgt_tokens
         self._mpn = MosesPunctNormalizer()
-        self._mpn.substitutions = [
-            (str(re.compile(r)), sub)
-            for r, sub in self._mpn.substitutions
-            if isinstance(r, str) and isinstance(sub, str)
+        self._mpn.substitutions = [  # type: ignore
+            (re.compile(r), sub) for r, sub in self._mpn.substitutions if isinstance(r, str) and isinstance(sub, str)
         ]
         self._stats = TrainStats()