Use new huggingface_hub tools for download models (#18438)

* Draft new cached_file

* Initial draft for config and model

* Small fixes

* Fix first batch of tests

* Look in cache when internet is down

* Fix last tests

* Bad black, not fixing all quality errors

* Make diff less

* Implement change for TF and Flax models

* Add tokenizer and feature extractor

* For compatibility with main

* Add utils to move the cache and auto-do it at first use.

* Quality

* Deal with empty commit shas

* Deal with empty etag

* Address review comments

tests/test_feature_extraction_common.py

@@ -170,13 +170,13 @@ class FeatureExtractorUtilTester(unittest.TestCase):
         # A mock response for an HTTP head request to emulate server down
         response_mock = mock.Mock()
         response_mock.status_code = 500
-        response_mock.headers = []
+        response_mock.headers = {}
         response_mock.raise_for_status.side_effect = HTTPError
 
         # Download this model to make sure it's in the cache.
         _ = Wav2Vec2FeatureExtractor.from_pretrained("hf-internal-testing/tiny-random-wav2vec2")
         # Under the mock environment we get a 500 error when trying to reach the model.
-        with mock.patch("transformers.utils.hub.requests.head", return_value=response_mock) as mock_head:
+        with mock.patch("requests.request", return_value=response_mock) as mock_head:
             _ = Wav2Vec2FeatureExtractor.from_pretrained("hf-internal-testing/tiny-random-wav2vec2")
             # This check we did call the fake head request
             mock_head.assert_called()