diff --git a/spacy/training/corpus.py b/spacy/training/corpus.py
index 37af9e476..c15c31a2a 100644
--- a/spacy/training/corpus.py
+++ b/spacy/training/corpus.py
@@ -6,7 +6,6 @@ from typing import TYPE_CHECKING, Callable, Iterable, Iterator, List, Optional,
 import srsly
 
 from .. import util
-from ..compat import Protocol
 from ..errors import Errors, Warnings
 from ..tokens import Doc, DocBin
 from ..vocab import Vocab
@@ -20,10 +19,6 @@ if TYPE_CHECKING:
 FILE_TYPE = ".spacy"
 
 
-class ReaderProtocol(Protocol):
-    def __call__(self, nlp: "Language") -> Iterable[Example]:
-        pass
-
 
 @util.registry.readers("spacy.Corpus.v1")
 def create_docbin_reader(
@@ -32,7 +27,7 @@ def create_docbin_reader(
     max_length: int = 0,
     limit: int = 0,
     augmenter: Optional[Callable] = None,
-) -> ReaderProtocol:
+) -> Callable[["Language"], Iterable[Example]]:
     if path is None:
         raise ValueError(Errors.E913)
     util.logger.debug("Loading corpus from path: %s", path)
@@ -51,7 +46,7 @@ def create_jsonl_reader(
     min_length: int = 0,
     max_length: int = 0,
     limit: int = 0,
-) -> ReaderProtocol:
+) -> Callable[["Language"], Iterable[Example]]:
     return JsonlCorpus(path, min_length=min_length, max_length=max_length, limit=limit)
 
 
@@ -69,7 +64,7 @@ def create_plain_text_reader(
     path: Optional[Path],
     min_length: int = 0,
     max_length: int = 0,
-) -> ReaderProtocol:
+) -> Callable[["Language"], Iterable[Example]]:
     """Iterate Example objects from a file or directory of plain text
     UTF-8 files with one line per doc.
 
@@ -150,7 +145,7 @@ class Corpus:
         self.augmenter = augmenter if augmenter is not None else dont_augment
         self.shuffle = shuffle
 
-    def __call__(self, nlp: "Language") -> Iterable[Example]:
+    def __call__(self, nlp: "Language") -> Iterator[Example]:
         """Yield examples from the data.
 
         nlp (Language): The current nlp object.
@@ -188,7 +183,7 @@ class Corpus:
 
     def make_examples(
         self, nlp: "Language", reference_docs: Iterable[Doc]
-    ) -> Iterable[Example]:
+    ) -> Iterator[Example]:
         for reference in reference_docs:
             if len(reference) == 0:
                 continue
@@ -203,7 +198,7 @@ class Corpus:
 
     def make_examples_gold_preproc(
         self, nlp: "Language", reference_docs: Iterable[Doc]
-    ) -> Iterable[Example]:
+    ) -> Iterator[Example]:
         for reference in reference_docs:
             if reference.has_annotation("SENT_START"):
                 ref_sents = [sent.as_doc() for sent in reference.sents]
@@ -263,7 +258,7 @@ class JsonlCorpus:
         self.max_length = max_length
         self.limit = limit
 
-    def __call__(self, nlp: "Language") -> Iterable[Example]:
+    def __call__(self, nlp: "Language") -> Iterator[Example]:
         """Yield examples from the data.
 
         nlp (Language): The current nlp object.
@@ -313,7 +308,7 @@ class PlainTextCorpus:
         self.min_length = min_length
         self.max_length = max_length
 
-    def __call__(self, nlp: "Language") -> Iterable[Example]:
+    def __call__(self, nlp: "Language") -> Iterator[Example]:
         """Yield examples from the data.
 
         nlp (Language): The current nlp object.