feat(api): allow converting inpaint models (#356)

api/onnx_web/convert/diffusion/diffusers.py

@@ -20,6 +20,7 @@ from diffusers import (
     AutoencoderKL,
     OnnxRuntimeModel,
     OnnxStableDiffusionPipeline,
+    StableDiffusionInpaintPipeline,
     StableDiffusionPipeline,
 )
 from onnx import load_model, save_model
@@ -183,6 +184,7 @@ def convert_diffusion_diffusers(
     source = source or model.get("source")
     single_vae = model.get("single_vae")
     replace_vae = model.get("vae")
+    pipe_type = model.get("pipeline", "image")
 
     device = conversion.training_device
     dtype = conversion.torch_dtype()
@@ -211,11 +213,21 @@ def convert_diffusion_diffusers(
             logger.info("ONNX model already exists, skipping")
             return (False, dest_path)
 
-    pipeline = StableDiffusionPipeline.from_pretrained(
-        source,
-        torch_dtype=dtype,
-        use_auth_token=conversion.token,
-    ).to(device)
+    if pipe_type == "image":
+        pipeline = StableDiffusionPipeline.from_pretrained(
+            source,
+            torch_dtype=dtype,
+            use_auth_token=conversion.token,
+        ).to(device)
+    elif pipe_type == "inpaint":
+        pipeline = StableDiffusionInpaintPipeline.from_pretrained(
+            source,
+            torch_dtype=dtype,
+            use_auth_token=conversion.token,
+        ).to(device)
+    else:
+        raise ValueError(f"unknown pipeline type: {pipe_type}")
+
     output_path = Path(dest_path)
 
     optimize_pipeline(conversion, pipeline)

api/schemas/extras.yaml

@@ -86,6 +86,12 @@ $defs:
             type: array
             items:
               $ref: "#/$defs/lora_network"
+          pipeline:
+            type: string
+            enum: [
+              image,
+              inpaint
+            ]
           vae:
             type: string