a few fix for SingleFile tests (#9522)

* update sd15 repo

* update more

Author: yiyixuxu

src/diffusers/loaders/single_file_utils.py

@@ -14,6 +14,7 @@
 # limitations under the License.
 """Conversion script for the Stable Diffusion checkpoints."""
 
+import copy
 import os
 import re
 from contextlib import nullcontext
@@ -91,11 +92,11 @@
     "xl_inpaint": {"pretrained_model_name_or_path": "diffusers/stable-diffusion-xl-1.0-inpainting-0.1"},
     "playground-v2-5": {"pretrained_model_name_or_path": "playgroundai/playground-v2.5-1024px-aesthetic"},
     "upscale": {"pretrained_model_name_or_path": "stabilityai/stable-diffusion-x4-upscaler"},
-    "inpainting": {"pretrained_model_name_or_path": "Lykon/dreamshaper-8-inpainting"},
+    "inpainting": {"pretrained_model_name_or_path": "stable-diffusion-v1-5/stable-diffusion-inpainting"},
     "inpainting_v2": {"pretrained_model_name_or_path": "stabilityai/stable-diffusion-2-inpainting"},
     "controlnet": {"pretrained_model_name_or_path": "lllyasviel/control_v11p_sd15_canny"},
     "v2": {"pretrained_model_name_or_path": "stabilityai/stable-diffusion-2-1"},
-    "v1": {"pretrained_model_name_or_path": "Lykon/dreamshaper-8"},
+    "v1": {"pretrained_model_name_or_path": "stable-diffusion-v1-5/stable-diffusion-v1-5"},
     "stable_cascade_stage_b": {"pretrained_model_name_or_path": "stabilityai/stable-cascade", "subfolder": "decoder"},
     "stable_cascade_stage_b_lite": {
         "pretrained_model_name_or_path": "stabilityai/stable-cascade",
@@ -541,6 +542,7 @@ def infer_diffusers_model_type(checkpoint):
 def fetch_diffusers_config(checkpoint):
     model_type = infer_diffusers_model_type(checkpoint)
     model_path = DIFFUSERS_DEFAULT_PIPELINE_PATHS[model_type]
+    model_path = copy.deepcopy(model_path)
 
     return model_path
 

tests/lora/test_lora_layers_sd.py

@@ -102,7 +102,7 @@ def tearDown(self):
     @slow
     @require_torch_gpu
     def test_integration_move_lora_cpu(self):
-        path = "Jiali/stable-diffusion-1.5"
+        path = "stable-diffusion-v1-5/stable-diffusion-v1-5"
         lora_id = "takuma104/lora-test-text-encoder-lora-target"
 
         pipe = StableDiffusionPipeline.from_pretrained(path, torch_dtype=torch.float16)
@@ -161,7 +161,7 @@ def test_integration_move_lora_cpu(self):
     def test_integration_move_lora_dora_cpu(self):
         from peft import LoraConfig
 
-        path = "Jiali/stable-diffusion-1.5"
+        path = "stable-diffusion-v1-5/stable-diffusion-v1-5"
         unet_lora_config = LoraConfig(
             init_lora_weights="gaussian",
             target_modules=["to_k", "to_q", "to_v", "to_out.0"],
@@ -221,7 +221,7 @@ def tearDown(self):
         torch.cuda.empty_cache()
 
     def test_integration_logits_with_scale(self):
-        path = "Jiali/stable-diffusion-1.5"
+        path = "stable-diffusion-v1-5/stable-diffusion-v1-5"
         lora_id = "takuma104/lora-test-text-encoder-lora-target"
 
         pipe = StableDiffusionPipeline.from_pretrained(path, torch_dtype=torch.float32)
@@ -253,7 +253,7 @@ def test_integration_logits_with_scale(self):
         release_memory(pipe)
 
     def test_integration_logits_no_scale(self):
-        path = "Jiali/stable-diffusion-1.5"
+        path = "stable-diffusion-v1-5/stable-diffusion-v1-5"
         lora_id = "takuma104/lora-test-text-encoder-lora-target"
 
         pipe = StableDiffusionPipeline.from_pretrained(path, torch_dtype=torch.float32)
@@ -284,7 +284,7 @@ def test_dreambooth_old_format(self):
 
         lora_model_id = "hf-internal-testing/lora_dreambooth_dog_example"
 
-        base_model_id = "Jiali/stable-diffusion-1.5"
+        base_model_id = "stable-diffusion-v1-5/stable-diffusion-v1-5"
 
         pipe = StableDiffusionPipeline.from_pretrained(base_model_id, safety_checker=None)
         pipe = pipe.to(torch_device)
@@ -308,7 +308,7 @@ def test_dreambooth_text_encoder_new_format(self):
 
         lora_model_id = "hf-internal-testing/lora-trained"
 
-        base_model_id = "Jiali/stable-diffusion-1.5"
+        base_model_id = "stable-diffusion-v1-5/stable-diffusion-v1-5"
 
         pipe = StableDiffusionPipeline.from_pretrained(base_model_id, safety_checker=None)
         pipe = pipe.to(torch_device)
@@ -419,9 +419,9 @@ def test_a1111_with_sequential_cpu_offload(self):
     def test_kohya_sd_v15_with_higher_dimensions(self):
         generator = torch.Generator().manual_seed(0)
 
-        pipe = StableDiffusionPipeline.from_pretrained("Jiali/stable-diffusion-1.5", safety_checker=None).to(
-            torch_device
-        )
+        pipe = StableDiffusionPipeline.from_pretrained(
+            "stable-diffusion-v1-5/stable-diffusion-v1-5", safety_checker=None
+        ).to(torch_device)
         lora_model_id = "hf-internal-testing/urushisato-lora"
         lora_filename = "urushisato_v15.safetensors"
         pipe.load_lora_weights(lora_model_id, weight_name=lora_filename)
@@ -444,7 +444,7 @@ def test_vanilla_funetuning(self):
 
         lora_model_id = "hf-internal-testing/sd-model-finetuned-lora-t4"
 
-        base_model_id = "Jiali/stable-diffusion-1.5"
+        base_model_id = "stable-diffusion-v1-5/stable-diffusion-v1-5"
 
         pipe = StableDiffusionPipeline.from_pretrained(base_model_id, safety_checker=None)
         pipe = pipe.to(torch_device)
@@ -467,9 +467,9 @@ def test_unload_kohya_lora(self):
         prompt = "masterpiece, best quality, mountain"
         num_inference_steps = 2
 
-        pipe = StableDiffusionPipeline.from_pretrained("Jiali/stable-diffusion-1.5", safety_checker=None).to(
-            torch_device
-        )
+        pipe = StableDiffusionPipeline.from_pretrained(
+            "stable-diffusion-v1-5/stable-diffusion-v1-5", safety_checker=None
+        ).to(torch_device)
         initial_images = pipe(
             prompt, output_type="np", generator=generator, num_inference_steps=num_inference_steps
         ).images
@@ -505,9 +505,9 @@ def test_load_unload_load_kohya_lora(self):
         prompt = "masterpiece, best quality, mountain"
         num_inference_steps = 2
 
-        pipe = StableDiffusionPipeline.from_pretrained("Jiali/stable-diffusion-1.5", safety_checker=None).to(
-            torch_device
-        )
+        pipe = StableDiffusionPipeline.from_pretrained(
+            "stable-diffusion-v1-5/stable-diffusion-v1-5", safety_checker=None
+        ).to(torch_device)
         initial_images = pipe(
             prompt, output_type="np", generator=generator, num_inference_steps=num_inference_steps
         ).images
@@ -547,9 +547,9 @@ def test_load_unload_load_kohya_lora(self):
 
     def test_not_empty_state_dict(self):
         # Makes sure https://github.com/huggingface/diffusers/issues/7054 does not happen again
-        pipe = AutoPipelineForText2Image.from_pretrained("Jiali/stable-diffusion-1.5", torch_dtype=torch.float16).to(
-            torch_device
-        )
+        pipe = AutoPipelineForText2Image.from_pretrained(
+            "stable-diffusion-v1-5/stable-diffusion-v1-5", torch_dtype=torch.float16
+        ).to(torch_device)
         pipe.scheduler = LCMScheduler.from_config(pipe.scheduler.config)
 
         cached_file = hf_hub_download("hf-internal-testing/lcm-lora-test-sd-v1-5", "test_lora.safetensors")
@@ -561,9 +561,9 @@ def test_not_empty_state_dict(self):
 
     def test_load_unload_load_state_dict(self):
         # Makes sure https://github.com/huggingface/diffusers/issues/7054 does not happen again
-        pipe = AutoPipelineForText2Image.from_pretrained("Jiali/stable-diffusion-1.5", torch_dtype=torch.float16).to(
-            torch_device
-        )
+        pipe = AutoPipelineForText2Image.from_pretrained(
+            "stable-diffusion-v1-5/stable-diffusion-v1-5", torch_dtype=torch.float16
+        ).to(torch_device)
         pipe.scheduler = LCMScheduler.from_config(pipe.scheduler.config)
 
         cached_file = hf_hub_download("hf-internal-testing/lcm-lora-test-sd-v1-5", "test_lora.safetensors")
@@ -580,7 +580,9 @@ def test_load_unload_load_state_dict(self):
         release_memory(pipe)
 
     def test_sdv1_5_lcm_lora(self):
-        pipe = DiffusionPipeline.from_pretrained("Jiali/stable-diffusion-1.5", torch_dtype=torch.float16)
+        pipe = DiffusionPipeline.from_pretrained(
+            "stable-diffusion-v1-5/stable-diffusion-v1-5", torch_dtype=torch.float16
+        )
         pipe.to(torch_device)
         pipe.scheduler = LCMScheduler.from_config(pipe.scheduler.config)
 
@@ -608,7 +610,9 @@ def test_sdv1_5_lcm_lora(self):
         release_memory(pipe)
 
     def test_sdv1_5_lcm_lora_img2img(self):
-        pipe = AutoPipelineForImage2Image.from_pretrained("Jiali/stable-diffusion-1.5", torch_dtype=torch.float16)
+        pipe = AutoPipelineForImage2Image.from_pretrained(
+            "stable-diffusion-v1-5/stable-diffusion-v1-5", torch_dtype=torch.float16
+        )
         pipe.to(torch_device)
         pipe.scheduler = LCMScheduler.from_config(pipe.scheduler.config)
 
@@ -649,7 +653,7 @@ def test_sd_load_civitai_empty_network_alpha(self):
         This test simply checks that loading a LoRA with an empty network alpha works fine
         See: https://github.com/huggingface/diffusers/issues/5606
         """
-        pipeline = StableDiffusionPipeline.from_pretrained("Jiali/stable-diffusion-1.5")
+        pipeline = StableDiffusionPipeline.from_pretrained("stable-diffusion-v1-5/stable-diffusion-v1-5")
         pipeline.enable_sequential_cpu_offload()
         civitai_path = hf_hub_download("ybelkada/test-ahi-civitai", "ahi_lora_weights.safetensors")
         pipeline.load_lora_weights(civitai_path, adapter_name="ahri")

tests/models/autoencoders/test_models_vae.py

@@ -1051,7 +1051,9 @@ def test_encode_decode(self):
 
     def test_sd(self):
         vae = ConsistencyDecoderVAE.from_pretrained("openai/consistency-decoder")  # TODO - update
-        pipe = StableDiffusionPipeline.from_pretrained("Jiali/stable-diffusion-1.5", vae=vae, safety_checker=None)
+        pipe = StableDiffusionPipeline.from_pretrained(
+            "stable-diffusion-v1-5/stable-diffusion-v1-5", vae=vae, safety_checker=None
+        )
         pipe.to(torch_device)
 
         out = pipe(
@@ -1099,7 +1101,7 @@ def test_sd_f16(self):
             "openai/consistency-decoder", torch_dtype=torch.float16
         )  # TODO - update
         pipe = StableDiffusionPipeline.from_pretrained(
-            "Jiali/stable-diffusion-1.5",
+            "stable-diffusion-v1-5/stable-diffusion-v1-5",
             torch_dtype=torch.float16,
             vae=vae,
             safety_checker=None,
@@ -1124,7 +1126,7 @@ def test_sd_f16(self):
     def test_vae_tiling(self):
         vae = ConsistencyDecoderVAE.from_pretrained("openai/consistency-decoder", torch_dtype=torch.float16)
         pipe = StableDiffusionPipeline.from_pretrained(
-            "Jiali/stable-diffusion-1.5", vae=vae, safety_checker=None, torch_dtype=torch.float16
+            "stable-diffusion-v1-5/stable-diffusion-v1-5", vae=vae, safety_checker=None, torch_dtype=torch.float16
         )
         pipe.to(torch_device)
         pipe.set_progress_bar_config(disable=None)

tests/pipelines/controlnet/test_controlnet.py

@@ -73,7 +73,7 @@ def _test_stable_diffusion_compile(in_queue, out_queue, timeout):
         controlnet = ControlNetModel.from_pretrained("lllyasviel/sd-controlnet-canny")
 
         pipe = StableDiffusionControlNetPipeline.from_pretrained(
-            "Jiali/stable-diffusion-1.5", safety_checker=None, controlnet=controlnet
+            "stable-diffusion-v1-5/stable-diffusion-v1-5", safety_checker=None, controlnet=controlnet
         )
         pipe.to("cuda")
         pipe.set_progress_bar_config(disable=None)
@@ -715,7 +715,7 @@ def test_canny(self):
         controlnet = ControlNetModel.from_pretrained("lllyasviel/sd-controlnet-canny")
 
         pipe = StableDiffusionControlNetPipeline.from_pretrained(
-            "Jiali/stable-diffusion-1.5", safety_checker=None, controlnet=controlnet
+            "stable-diffusion-v1-5/stable-diffusion-v1-5", safety_checker=None, controlnet=controlnet
         )
         pipe.enable_model_cpu_offload()
         pipe.set_progress_bar_config(disable=None)
@@ -742,7 +742,7 @@ def test_depth(self):
         controlnet = ControlNetModel.from_pretrained("lllyasviel/sd-controlnet-depth")
 
         pipe = StableDiffusionControlNetPipeline.from_pretrained(
-            "Jiali/stable-diffusion-1.5", safety_checker=None, controlnet=controlnet
+            "stable-diffusion-v1-5/stable-diffusion-v1-5", safety_checker=None, controlnet=controlnet
         )
         pipe.enable_model_cpu_offload()
         pipe.set_progress_bar_config(disable=None)
@@ -769,7 +769,7 @@ def test_hed(self):
         controlnet = ControlNetModel.from_pretrained("lllyasviel/sd-controlnet-hed")
 
         pipe = StableDiffusionControlNetPipeline.from_pretrained(
-            "Jiali/stable-diffusion-1.5", safety_checker=None, controlnet=controlnet
+            "stable-diffusion-v1-5/stable-diffusion-v1-5", safety_checker=None, controlnet=controlnet
         )
         pipe.enable_model_cpu_offload()
         pipe.set_progress_bar_config(disable=None)
@@ -796,7 +796,7 @@ def test_mlsd(self):
         controlnet = ControlNetModel.from_pretrained("lllyasviel/sd-controlnet-mlsd")
 
         pipe = StableDiffusionControlNetPipeline.from_pretrained(
-            "Jiali/stable-diffusion-1.5", safety_checker=None, controlnet=controlnet
+            "stable-diffusion-v1-5/stable-diffusion-v1-5", safety_checker=None, controlnet=controlnet
         )
         pipe.enable_model_cpu_offload()
         pipe.set_progress_bar_config(disable=None)
@@ -823,7 +823,7 @@ def test_normal(self):
         controlnet = ControlNetModel.from_pretrained("lllyasviel/sd-controlnet-normal")
 
         pipe = StableDiffusionControlNetPipeline.from_pretrained(
-            "Jiali/stable-diffusion-1.5", safety_checker=None, controlnet=controlnet
+            "stable-diffusion-v1-5/stable-diffusion-v1-5", safety_checker=None, controlnet=controlnet
         )
         pipe.enable_model_cpu_offload()
         pipe.set_progress_bar_config(disable=None)
@@ -850,7 +850,7 @@ def test_openpose(self):
         controlnet = ControlNetModel.from_pretrained("lllyasviel/sd-controlnet-openpose")
 
         pipe = StableDiffusionControlNetPipeline.from_pretrained(
-            "Jiali/stable-diffusion-1.5", safety_checker=None, controlnet=controlnet
+            "stable-diffusion-v1-5/stable-diffusion-v1-5", safety_checker=None, controlnet=controlnet
         )
         pipe.enable_model_cpu_offload()
         pipe.set_progress_bar_config(disable=None)
@@ -877,7 +877,7 @@ def test_scribble(self):
         controlnet = ControlNetModel.from_pretrained("lllyasviel/sd-controlnet-scribble")
 
         pipe = StableDiffusionControlNetPipeline.from_pretrained(
-            "Jiali/stable-diffusion-1.5", safety_checker=None, controlnet=controlnet
+            "stable-diffusion-v1-5/stable-diffusion-v1-5", safety_checker=None, controlnet=controlnet
         )
         pipe.enable_model_cpu_offload()
         pipe.set_progress_bar_config(disable=None)
@@ -904,7 +904,7 @@ def test_seg(self):
         controlnet = ControlNetModel.from_pretrained("lllyasviel/sd-controlnet-seg")
 
         pipe = StableDiffusionControlNetPipeline.from_pretrained(
-            "Jiali/stable-diffusion-1.5", safety_checker=None, controlnet=controlnet
+            "stable-diffusion-v1-5/stable-diffusion-v1-5", safety_checker=None, controlnet=controlnet
         )
         pipe.enable_model_cpu_offload()
         pipe.set_progress_bar_config(disable=None)
@@ -935,7 +935,7 @@ def test_sequential_cpu_offloading(self):
         controlnet = ControlNetModel.from_pretrained("lllyasviel/sd-controlnet-seg")
 
         pipe = StableDiffusionControlNetPipeline.from_pretrained(
-            "Jiali/stable-diffusion-1.5", safety_checker=None, controlnet=controlnet
+            "stable-diffusion-v1-5/stable-diffusion-v1-5", safety_checker=None, controlnet=controlnet
         )
         pipe.set_progress_bar_config(disable=None)
         pipe.enable_attention_slicing()
@@ -961,7 +961,7 @@ def test_canny_guess_mode(self):
         controlnet = ControlNetModel.from_pretrained("lllyasviel/sd-controlnet-canny")
 
         pipe = StableDiffusionControlNetPipeline.from_pretrained(
-            "Jiali/stable-diffusion-1.5", safety_checker=None, controlnet=controlnet
+            "stable-diffusion-v1-5/stable-diffusion-v1-5", safety_checker=None, controlnet=controlnet
         )
         pipe.enable_model_cpu_offload()
         pipe.set_progress_bar_config(disable=None)
@@ -993,7 +993,7 @@ def test_canny_guess_mode_euler(self):
         controlnet = ControlNetModel.from_pretrained("lllyasviel/sd-controlnet-canny")
 
         pipe = StableDiffusionControlNetPipeline.from_pretrained(
-            "Jiali/stable-diffusion-1.5", safety_checker=None, controlnet=controlnet
+            "stable-diffusion-v1-5/stable-diffusion-v1-5", safety_checker=None, controlnet=controlnet
         )
         pipe.scheduler = EulerDiscreteScheduler.from_config(pipe.scheduler.config)
         pipe.enable_model_cpu_offload()
@@ -1035,7 +1035,7 @@ def test_v11_shuffle_global_pool_conditions(self):
         controlnet = ControlNetModel.from_pretrained("lllyasviel/control_v11e_sd15_shuffle")
 
         pipe = StableDiffusionControlNetPipeline.from_pretrained(
-            "Jiali/stable-diffusion-1.5", safety_checker=None, controlnet=controlnet
+            "stable-diffusion-v1-5/stable-diffusion-v1-5", safety_checker=None, controlnet=controlnet
         )
         pipe.enable_model_cpu_offload()
         pipe.set_progress_bar_config(disable=None)
@@ -1081,7 +1081,9 @@ def test_pose_and_canny(self):
         controlnet_pose = ControlNetModel.from_pretrained("lllyasviel/sd-controlnet-openpose")
 
         pipe = StableDiffusionControlNetPipeline.from_pretrained(
-            "Jiali/stable-diffusion-1.5", safety_checker=None, controlnet=[controlnet_pose, controlnet_canny]
+            "stable-diffusion-v1-5/stable-diffusion-v1-5",
+            safety_checker=None,
+            controlnet=[controlnet_pose, controlnet_canny],
         )
         pipe.enable_model_cpu_offload()
         pipe.set_progress_bar_config(disable=None)

tests/pipelines/controlnet/test_controlnet_img2img.py

@@ -407,7 +407,7 @@ def test_canny(self):
         controlnet = ControlNetModel.from_pretrained("lllyasviel/sd-controlnet-canny")
 
         pipe = StableDiffusionControlNetImg2ImgPipeline.from_pretrained(
-            "Jiali/stable-diffusion-1.5", safety_checker=None, controlnet=controlnet
+            "stable-diffusion-v1-5/stable-diffusion-v1-5", safety_checker=None, controlnet=controlnet
         )
         pipe.enable_model_cpu_offload()
         pipe.set_progress_bar_config(disable=None)

tests/pipelines/controlnet/test_controlnet_inpaint.py

@@ -504,7 +504,7 @@ def test_inpaint(self):
         controlnet = ControlNetModel.from_pretrained("lllyasviel/control_v11p_sd15_inpaint")
 
         pipe = StableDiffusionControlNetInpaintPipeline.from_pretrained(
-            "Jiali/stable-diffusion-1.5", safety_checker=None, controlnet=controlnet
+            "stable-diffusion-v1-5/stable-diffusion-v1-5", safety_checker=None, controlnet=controlnet
         )
         pipe.scheduler = DDIMScheduler.from_config(pipe.scheduler.config)
         pipe.enable_model_cpu_offload()

tests/pipelines/controlnet/test_flax_controlnet.py

@@ -41,7 +41,7 @@ def test_canny(self):
             "lllyasviel/sd-controlnet-canny", from_pt=True, dtype=jnp.bfloat16
         )
         pipe, params = FlaxStableDiffusionControlNetPipeline.from_pretrained(
-            "Jiali/stable-diffusion-1.5", controlnet=controlnet, from_pt=True, dtype=jnp.bfloat16
+            "stable-diffusion-v1-5/stable-diffusion-v1-5", controlnet=controlnet, from_pt=True, dtype=jnp.bfloat16
         )
         params["controlnet"] = controlnet_params
 
@@ -86,7 +86,7 @@ def test_pose(self):
             "lllyasviel/sd-controlnet-openpose", from_pt=True, dtype=jnp.bfloat16
         )
         pipe, params = FlaxStableDiffusionControlNetPipeline.from_pretrained(
-            "Jiali/stable-diffusion-1.5", controlnet=controlnet, from_pt=True, dtype=jnp.bfloat16
+            "stable-diffusion-v1-5/stable-diffusion-v1-5", controlnet=controlnet, from_pt=True, dtype=jnp.bfloat16
         )
         params["controlnet"] = controlnet_params