hopefully ,final?

Author: sayakpaul

tests/pipelines/kandinsky3/test_kandinsky3.py

@@ -109,7 +109,7 @@ def get_dummy_components(self, time_cond_proj_dim=None):
         movq = self.dummy_movq
         torch.manual_seed(0)
         config = AutoConfig.from_pretrained("hf-internal-testing/tiny-random-t5")
-        text_encoder = T5EncoderModel(config)
+        text_encoder = T5EncoderModel(config).eval()
 
         torch.manual_seed(0)
         tokenizer = AutoTokenizer.from_pretrained("hf-internal-testing/tiny-random-t5")
@@ -156,7 +156,7 @@ def test_kandinsky3(self):
 
         assert image.shape == (1, 16, 16, 3)
 
-        expected_slice = np.array([0.3768, 0.4373, 0.4865, 0.4890, 0.4299, 0.5122, 0.4921, 0.4924, 0.5599])
+        expected_slice = np.array([0.3301, 0.3106, 0.4827, 0.5204, 0.4242, 0.4987, 0.4665, 0.5120, 0.5558])
 
         assert np.abs(image_slice.flatten() - expected_slice).max() < 1e-2, (
             f" expected_slice {expected_slice}, but got {image_slice.flatten()}"

tests/pipelines/kandinsky3/test_kandinsky3_img2img.py

@@ -120,7 +120,7 @@ def get_dummy_components(self, time_cond_proj_dim=None):
         movq = self.dummy_movq
         torch.manual_seed(0)
         config = AutoConfig.from_pretrained("hf-internal-testing/tiny-random-t5")
-        text_encoder = T5EncoderModel(config)
+        text_encoder = T5EncoderModel(config).eval()
 
         torch.manual_seed(0)
         tokenizer = AutoTokenizer.from_pretrained("hf-internal-testing/tiny-random-t5")
@@ -158,7 +158,7 @@ def get_dummy_inputs(self, device, seed=0):
     def test_dict_tuple_outputs_equivalent(self):
         expected_slice = None
         if torch_device == "cpu":
-            expected_slice = np.array([0.5762, 0.6112, 0.4150, 0.6018, 0.6167, 0.4626, 0.5426, 0.5641, 0.6536])
+            expected_slice = np.array([0.5261, 0.5688, 0.4093, 0.4865, 0.5326, 0.4480, 0.5064, 0.5113, 0.6222])
         super().test_dict_tuple_outputs_equivalent(expected_slice=expected_slice)
 
     def test_kandinsky3_img2img(self):
@@ -178,9 +178,7 @@ def test_kandinsky3_img2img(self):
 
         assert image.shape == (1, 64, 64, 3)
 
-        expected_slice = np.array(
-            [0.576259, 0.6132097, 0.41703486, 0.603196, 0.62062526, 0.4655338, 0.5434324, 0.5660727, 0.65433365]
-        )
+        expected_slice = np.array([0.5261, 0.5688, 0.4093, 0.4865, 0.5326, 0.4480, 0.5064, 0.5113, 0.6222])
 
         assert np.abs(image_slice.flatten() - expected_slice).max() < 1e-2, (
             f" expected_slice {expected_slice}, but got {image_slice.flatten()}"

tests/pipelines/qwenimage/test_qwenimage.py

@@ -113,7 +113,7 @@ def get_dummy_components(self):
             vision_start_token_id=151652,
             vision_token_id=151654,
         )
-        text_encoder = Qwen2_5_VLForConditionalGeneration(config)
+        text_encoder = Qwen2_5_VLForConditionalGeneration(config).eval()
         tokenizer = Qwen2Tokenizer.from_pretrained("hf-internal-testing/tiny-random-Qwen2VLForConditionalGeneration")
 
         components = {
@@ -160,7 +160,7 @@ def test_inference(self):
         self.assertEqual(generated_image.shape, (3, 32, 32))
 
         # fmt: off
-        expected_slice = torch.tensor([0.5646, 0.6369, 0.6019, 0.5640, 0.5830, 0.5520, 0.5717, 0.6315, 0.4167, 0.3563, 0.5640, 0.4849, 0.4961, 0.5237, 0.4084, 0.5014])
+        expected_slice = torch.tensor([0.5633, 0.6368, 0.6015, 0.5637, 0.5817, 0.5528, 0.5718, 0.6326, 0.4147, 0.3556, 0.5623, 0.4833, 0.4971, 0.5262, 0.4087, 0.5021])
         # fmt: on
 
         generated_slice = generated_image.flatten()

tests/pipelines/qwenimage/test_qwenimage_edit.py

@@ -115,7 +115,7 @@ def get_dummy_components(self):
             vision_start_token_id=151652,
             vision_token_id=151654,
         )
-        text_encoder = Qwen2_5_VLForConditionalGeneration(config)
+        text_encoder = Qwen2_5_VLForConditionalGeneration(config).eval()
         tokenizer = Qwen2Tokenizer.from_pretrained(tiny_ckpt_id)
 
         components = {
@@ -163,7 +163,7 @@ def test_inference(self):
         self.assertEqual(generated_image.shape, (3, 32, 32))
 
         # fmt: off
-        expected_slice = torch.tensor([0.5640, 0.6350, 0.6003, 0.5606, 0.5801, 0.5502, 0.5757, 0.6388, 0.4174, 0.3590, 0.5647, 0.4891, 0.4975, 0.5256, 0.4088, 0.4991])
+        expected_slice = torch.tensor([0.5637, 0.6341, 0.6001, 0.5620, 0.5794, 0.5498, 0.5757, 0.6389, 0.4174, 0.3597, 0.5649, 0.4894, 0.4969, 0.5255, 0.4083, 0.4986])
         # fmt: on
 
         generated_slice = generated_image.flatten()