Update conversion script to correctly handle SD 2 (huggingface#1511)

* Conversion SD 2

* finish

Author: patrickvonplaten

scripts/convert_original_stable_diffusion_to_diffusers.py

@@ -33,6 +33,7 @@
     DPMSolverMultistepScheduler,
     EulerAncestralDiscreteScheduler,
     EulerDiscreteScheduler,
+    HeunDiscreteScheduler,
     LDMTextToImagePipeline,
     LMSDiscreteScheduler,
     PNDMScheduler,
@@ -232,6 +233,15 @@ def create_unet_diffusers_config(original_config, image_size: int):
 
     vae_scale_factor = 2 ** (len(vae_params.ch_mult) - 1)
 
+    head_dim = unet_params.num_heads if "num_heads" in unet_params else None
+    use_linear_projection = (
+        unet_params.use_linear_in_transformer if "use_linear_in_transformer" in unet_params else False
+    )
+    if use_linear_projection:
+        # stable diffusion 2-base-512 and 2-768
+        if head_dim is None:
+            head_dim = [5, 10, 20, 20]
+
     config = dict(
         sample_size=image_size // vae_scale_factor,
         in_channels=unet_params.in_channels,
@@ -241,7 +251,8 @@ def create_unet_diffusers_config(original_config, image_size: int):
         block_out_channels=tuple(block_out_channels),
         layers_per_block=unet_params.num_res_blocks,
         cross_attention_dim=unet_params.context_dim,
-        attention_head_dim=unet_params.num_heads,
+        attention_head_dim=head_dim,
+        use_linear_projection=use_linear_projection,
     )
 
     return config
@@ -636,6 +647,22 @@ def convert_ldm_clip_checkpoint(checkpoint):
     return text_model
 
 
+def convert_open_clip_checkpoint(checkpoint):
+    text_model = CLIPTextModel.from_pretrained("stabilityai/stable-diffusion-2", subfolder="text_encoder")
+
+    # SKIP for now - need openclip -> HF conversion script here
+    #    keys = list(checkpoint.keys())
+    #
+    #    text_model_dict = {}
+    #    for key in keys:
+    #        if key.startswith("cond_stage_model.model.transformer"):
+    #            text_model_dict[key[len("cond_stage_model.model.transformer.") :]] = checkpoint[key]
+    #
+    #    text_model.load_state_dict(text_model_dict)
+
+    return text_model
+
+
 if __name__ == "__main__":
     parser = argparse.ArgumentParser()
 
@@ -657,13 +684,22 @@ def convert_ldm_clip_checkpoint(checkpoint):
     )
     parser.add_argument(
         "--image_size",
-        default=512,
+        default=None,
         type=int,
         help=(
             "The image size that the model was trained on. Use 512 for Stable Diffusion v1.X and Stable Siffusion v2"
             " Base. Use 768 for Stable Diffusion v2."
         ),
     )
+    parser.add_argument(
+        "--prediction_type",
+        default=None,
+        type=int,
+        help=(
+            "The prediction type that the model was trained on. Use 'epsilon' for Stable Diffusion v1.X and Stable"
+            " Siffusion v2 Base. Use 'v-prediction' for Stable Diffusion v2."
+        ),
+    )
     parser.add_argument(
         "--extract_ema",
         action="store_true",
@@ -674,73 +710,117 @@ def convert_ldm_clip_checkpoint(checkpoint):
         ),
     )
     parser.add_argument("--dump_path", default=None, type=str, required=True, help="Path to the output model.")
-
     args = parser.parse_args()
 
+    image_size = args.image_size
+    prediction_type = args.prediction_type
+
+    checkpoint = torch.load(args.checkpoint_path)
+    global_step = checkpoint["global_step"]
+    checkpoint = checkpoint["state_dict"]
+
     if args.original_config_file is None:
-        os.system(
-            "wget https://raw.githubusercontent.com/CompVis/stable-diffusion/main/configs/stable-diffusion/v1-inference.yaml"
-        )
-        args.original_config_file = "./v1-inference.yaml"
+        key_name = "model.diffusion_model.input_blocks.2.1.transformer_blocks.0.attn2.to_k.weight"
+
+        if key_name in checkpoint and checkpoint[key_name].shape[-1] == 1024:
+            # model_type = "v2"
+            os.system(
+                "wget https://raw.githubusercontent.com/Stability-AI/stablediffusion/main/configs/stable-diffusion/v2-inference-v.yaml"
+            )
+            args.original_config_file = "./v2-inference-v.yaml"
+        else:
+            # model_type = "v1"
+            os.system(
+                "wget https://raw.githubusercontent.com/CompVis/stable-diffusion/main/configs/stable-diffusion/v1-inference.yaml"
+            )
+            args.original_config_file = "./v1-inference.yaml"
 
     original_config = OmegaConf.load(args.original_config_file)
 
-    checkpoint = torch.load(args.checkpoint_path)
-    checkpoint = checkpoint["state_dict"]
+    if (
+        "parameterization" in original_config["model"]["params"]
+        and original_config["model"]["params"]["parameterization"] == "v"
+    ):
+        if prediction_type is None:
+            # NOTE: For stable diffusion 2 base it is recommended to pass `prediction_type=="epsilon"`
+            # as it relies on a brittle global step parameter here
+            prediction_type = "epsilon" if global_step == 875000 else "v_prediction"
+        if image_size is None:
+            # NOTE: For stable diffusion 2 base one has to pass `image_size==512`
+            # as it relies on a brittle global step parameter here
+            image_size = 512 if global_step == 875000 else 768
+    else:
+        if prediction_type is None:
+            prediction_type = "epsilon"
+        if image_size is None:
+            image_size = 512
 
     num_train_timesteps = original_config.model.params.timesteps
     beta_start = original_config.model.params.linear_start
     beta_end = original_config.model.params.linear_end
+
+    scheduler = DDIMScheduler(
+        beta_end=beta_end,
+        beta_schedule="scaled_linear",
+        beta_start=beta_start,
+        num_train_timesteps=num_train_timesteps,
+        steps_offset=1,
+        clip_sample=False,
+        set_alpha_to_one=False,
+        prediction_type=prediction_type,
+    )
     if args.scheduler_type == "pndm":
-        scheduler = PNDMScheduler(
-            beta_end=beta_end,
-            beta_schedule="scaled_linear",
-            beta_start=beta_start,
-            num_train_timesteps=num_train_timesteps,
-            skip_prk_steps=True,
-        )
+        config = dict(scheduler.config)
+        config["skip_prk_steps"] = True
+        scheduler = PNDMScheduler.from_config(config)
     elif args.scheduler_type == "lms":
-        scheduler = LMSDiscreteScheduler(beta_start=beta_start, beta_end=beta_end, beta_schedule="scaled_linear")
+        scheduler = LMSDiscreteScheduler.from_config(scheduler.config)
+    elif args.scheduler_type == "heun":
+        scheduler = HeunDiscreteScheduler.from_config(scheduler.config)
     elif args.scheduler_type == "euler":
-        scheduler = EulerDiscreteScheduler(beta_start=beta_start, beta_end=beta_end, beta_schedule="scaled_linear")
+        scheduler = EulerDiscreteScheduler.from_config(scheduler.config)
     elif args.scheduler_type == "euler-ancestral":
-        scheduler = EulerAncestralDiscreteScheduler(
-            beta_start=beta_start, beta_end=beta_end, beta_schedule="scaled_linear"
-        )
+        scheduler = EulerAncestralDiscreteScheduler.from_config(scheduler.config)
     elif args.scheduler_type == "dpm":
-        scheduler = DPMSolverMultistepScheduler(
-            beta_start=beta_start, beta_end=beta_end, beta_schedule="scaled_linear"
-        )
+        scheduler = DPMSolverMultistepScheduler.from_config(scheduler.config)
     elif args.scheduler_type == "ddim":
-        scheduler = DDIMScheduler(
-            beta_start=beta_start,
-            beta_end=beta_end,
-            beta_schedule="scaled_linear",
-            clip_sample=False,
-            set_alpha_to_one=False,
-        )
+        scheduler = scheduler
     else:
         raise ValueError(f"Scheduler of type {args.scheduler_type} doesn't exist!")
 
     # Convert the UNet2DConditionModel model.
-    unet_config = create_unet_diffusers_config(original_config, image_size=args.image_size)
+    unet_config = create_unet_diffusers_config(original_config, image_size=image_size)
+    unet = UNet2DConditionModel(**unet_config)
+
     converted_unet_checkpoint = convert_ldm_unet_checkpoint(
         checkpoint, unet_config, path=args.checkpoint_path, extract_ema=args.extract_ema
     )
 
-    unet = UNet2DConditionModel(**unet_config)
     unet.load_state_dict(converted_unet_checkpoint)
 
     # Convert the VAE model.
-    vae_config = create_vae_diffusers_config(original_config, image_size=args.image_size)
+    vae_config = create_vae_diffusers_config(original_config, image_size=image_size)
     converted_vae_checkpoint = convert_ldm_vae_checkpoint(checkpoint, vae_config)
 
     vae = AutoencoderKL(**vae_config)
     vae.load_state_dict(converted_vae_checkpoint)
 
     # Convert the text model.
     text_model_type = original_config.model.params.cond_stage_config.target.split(".")[-1]
-    if text_model_type == "FrozenCLIPEmbedder":
+    if text_model_type == "FrozenOpenCLIPEmbedder":
+        text_model = convert_open_clip_checkpoint(checkpoint)
+        tokenizer = CLIPTokenizer.from_pretrained("stabilityai/stable-diffusion-2", subfolder="tokenizer")
+        pipe = StableDiffusionPipeline(
+            vae=vae,
+            text_encoder=text_model,
+            tokenizer=tokenizer,
+            unet=unet,
+            scheduler=scheduler,
+            safety_checker=None,
+            feature_extractor=None,
+            requires_safety_checker=False,
+        )
+    elif text_model_type == "FrozenCLIPEmbedder":
         text_model = convert_ldm_clip_checkpoint(checkpoint)
         tokenizer = CLIPTokenizer.from_pretrained("openai/clip-vit-large-patch14")
         safety_checker = StableDiffusionSafetyChecker.from_pretrained("CompVis/stable-diffusion-safety-checker")