From f4809307e409d5ce698364ad48b69d38e0c406e9 Mon Sep 17 00:00:00 2001
From: fxmarty <9808326+fxmarty@users.noreply.github.com>
Date: Fri, 14 Jun 2024 16:33:33 +0200
Subject: [PATCH] Remove read token (#1903)

* remove read token

* rename var & use org model

* style & remove token

* fix failing tests on datasets release
---
 .github/workflows/test_onnxruntime.yml |  2 ++
 optimum/utils/testing_utils.py         |  3 ---
 tests/onnxruntime/test_modeling.py     | 11 +++++++----
 tests/utils/test_task_processors.py    |  7 ++++++-
 4 files changed, 15 insertions(+), 8 deletions(-)

diff --git a/.github/workflows/test_onnxruntime.yml b/.github/workflows/test_onnxruntime.yml
index 4893b681a66..291a3b08335 100644
--- a/.github/workflows/test_onnxruntime.yml
+++ b/.github/workflows/test_onnxruntime.yml
@@ -50,6 +50,8 @@ jobs:
           pytest onnxruntime -m "run_in_series" --durations=0 -vvvv -s
 
       - name: Test with pytest (in parallel)
+        env:
+          FXMARTYCLONE_READ_TOKEN: ${{ secrets.HF_HUB_READ_TOKEN }}
         working-directory: tests
         run: |
           pytest onnxruntime -m "not run_in_series" --durations=0 -vvvv -s -n auto
diff --git a/optimum/utils/testing_utils.py b/optimum/utils/testing_utils.py
index f1c2f668e3c..a7c2b8bb050 100644
--- a/optimum/utils/testing_utils.py
+++ b/optimum/utils/testing_utils.py
@@ -36,9 +36,6 @@
 # Used to test the hub
 USER = "__DUMMY_OPTIMUM_USER__"
 
-# Not critical, only usable on the sandboxed CI instance.
-TOKEN = "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"
-
 
 def flatten_dict(dictionary: Dict):
     """
diff --git a/tests/onnxruntime/test_modeling.py b/tests/onnxruntime/test_modeling.py
index 3fe2c5e14dc..7b2c8a66b9e 100644
--- a/tests/onnxruntime/test_modeling.py
+++ b/tests/onnxruntime/test_modeling.py
@@ -938,11 +938,14 @@ def test_stable_diffusion_model_on_rocm_ep_str(self):
         self.assertListEqual(model.providers, ["ROCMExecutionProvider", "CPUExecutionProvider"])
 
     def test_load_model_from_hub_private(self):
-        subprocess.run("huggingface-cli logout", shell=True)
-        # Read token of fxmartyclone (dummy user).
-        token = "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"
+        token = os.environ.get("HF_HUB_READ_TOKEN", None)
 
-        model = ORTModelForCustomTasks.from_pretrained("fxmartyclone/tiny-onnx-private-2", use_auth_token=token)
+        if token is None:
+            self.skipTest("Test requires a token for fxmartyclone in the environment variable `HF_HUB_READ_TOKEN`.")
+
+        model = ORTModelForCustomTasks.from_pretrained(
+            "optimum-internal-testing/tiny-random-phi-private", use_auth_token=token
+        )
         self.assertIsInstance(model.model, onnxruntime.InferenceSession)
         self.assertIsInstance(model.config, PretrainedConfig)
 
diff --git a/tests/utils/test_task_processors.py b/tests/utils/test_task_processors.py
index af89aec2b90..16567048073 100644
--- a/tests/utils/test_task_processors.py
+++ b/tests/utils/test_task_processors.py
@@ -50,7 +50,7 @@
         "dataset_data_keys": {"question": "question", "context": "answer"},
     },
     "image-classification": {
-        "dataset_args": "mnist",
+        "dataset_args": "sasha/dog-food",
         "dataset_data_keys": {"image": "image"},
     },
 }
@@ -232,6 +232,11 @@ def test_load_dataset_with_max_length(self):
         input_ids = dataset[0]["input_ids"]
         self.assertEqual(len(input_ids), max_length)
 
+    def test_load_default_dataset(self):
+        self.skipTest(
+            "Skipping so as not to execute conll2003 remote code (test would require trust_remote_code=True)"
+        )
+
 
 class QuestionAnsweringProcessorTest(TestCase, TaskProcessorTestBase):
     TASK_NAME = "question-answering"