Upload folder using huggingface_hub

.gitattributes

@@ -2,7 +2,6 @@
 *.arrow filter=lfs diff=lfs merge=lfs -text
 *.bin filter=lfs diff=lfs merge=lfs -text
 *.bz2 filter=lfs diff=lfs merge=lfs -text
-*.ckpt filter=lfs diff=lfs merge=lfs -text
 *.ftz filter=lfs diff=lfs merge=lfs -text
 *.gz filter=lfs diff=lfs merge=lfs -text
 *.h5 filter=lfs diff=lfs merge=lfs -text
@@ -22,10 +21,8 @@
 *.pt filter=lfs diff=lfs merge=lfs -text
 *.pth filter=lfs diff=lfs merge=lfs -text
 *.rar filter=lfs diff=lfs merge=lfs -text
-*.safetensors filter=lfs diff=lfs merge=lfs -text
 saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.tar.* filter=lfs diff=lfs merge=lfs -text
-*.tar filter=lfs diff=lfs merge=lfs -text
 *.tflite filter=lfs diff=lfs merge=lfs -text
 *.tgz filter=lfs diff=lfs merge=lfs -text
 *.wasm filter=lfs diff=lfs merge=lfs -text
@@ -33,3 +30,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+diffusion_pytorch_model.safetensors filter=lfs diff=lfs merge=lfs -text

config.json

@@ -0,0 +1,37 @@
+{
+  "_class_name": "AutoencoderKL",
+  "_diffusers_version": "0.30.0.dev0",
+  "act_fn": "silu",
+  "block_out_channels": [
+    128,
+    256,
+    512,
+    512
+  ],
+  "down_block_types": [
+    "DownEncoderBlock2D",
+    "DownEncoderBlock2D",
+    "DownEncoderBlock2D",
+    "DownEncoderBlock2D"
+  ],
+  "force_upcast": true,
+  "in_channels": 3,
+  "latent_channels": 16,
+  "latents_mean": null,
+  "latents_std": null,
+  "layers_per_block": 2,
+  "mid_block_add_attention": true,
+  "norm_num_groups": 32,
+  "out_channels": 3,
+  "sample_size": 1024,
+  "scaling_factor": 0.3611,
+  "shift_factor": 0.1159,
+  "up_block_types": [
+    "UpDecoderBlock2D",
+    "UpDecoderBlock2D",
+    "UpDecoderBlock2D",
+    "UpDecoderBlock2D"
+  ],
+  "use_post_quant_conv": false,
+  "use_quant_conv": false
+}

diffusion_pytorch_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f5b59a26851551b67ae1fe58d32e76486e1e812def4696a4bea97f16604d40a3
+size 167666902

handler.py

@@ -0,0 +1,81 @@
+from typing import Dict, List, Any
+import torch
+from base64 import b64decode
+from diffusers import AutoencoderKL
+from diffusers.image_processor import VaeImageProcessor
+
+
+class EndpointHandler:
+    def __init__(self, path=""):
+        self.device = "cuda"
+        self.dtype = torch.bfloat16
+        self.vae = (
+            AutoencoderKL.from_pretrained(path, torch_dtype=self.dtype)
+            .to(self.device, self.dtype)
+            .eval()
+        )
+
+        self.vae_scale_factor = 2 ** (len(self.vae.config.block_out_channels) - 1)
+        self.image_processor = VaeImageProcessor(vae_scale_factor=self.vae_scale_factor)
+
+    @staticmethod
+    def _unpack_latents(latents, height, width, vae_scale_factor):
+        batch_size, num_patches, channels = latents.shape
+
+        # VAE applies 8x compression on images but we must also account for packing which requires
+        # latent height and width to be divisible by 2.
+        height = 2 * (int(height) // (vae_scale_factor * 2))
+        width = 2 * (int(width) // (vae_scale_factor * 2))
+
+        latents = latents.view(batch_size, height // 2, width // 2, channels // 4, 2, 2)
+        latents = latents.permute(0, 3, 1, 4, 2, 5)
+
+        latents = latents.reshape(batch_size, channels // (2 * 2), height, width)
+
+        return latents
+
+    @torch.no_grad()
+    def __call__(self, data: Any) -> List[List[Dict[str, float]]]:
+        """
+        Args:
+            data (:obj:):
+                includes the input data and the parameters for the inference.
+        """
+        tensor = data["inputs"]
+        tensor = b64decode(tensor.encode("utf-8"))
+        parameters = data.get("parameters", {})
+        if "shape" not in parameters:
+            raise ValueError("Expected `shape` in parameters.")
+        if "dtype" not in parameters:
+            raise ValueError("Expected `dtype` in parameters.")
+        if "height" not in parameters:
+            raise ValueError("Expected `height` in parameters.")
+        if "width" not in parameters:
+            raise ValueError("Expected `width` in parameters.")
+
+        DTYPE_MAP = {
+            "float16": torch.float16,
+            "float32": torch.float32,
+            "bfloat16": torch.bfloat16,
+        }
+
+        shape = parameters.get("shape")
+        dtype = DTYPE_MAP.get(parameters.get("dtype"))
+        height = parameters.get("height")
+        width = parameters.get("width")
+
+        tensor = torch.frombuffer(bytearray(tensor), dtype=dtype).reshape(shape)
+
+        tensor = tensor.to(self.device, self.dtype)
+
+        tensor = self._unpack_latents(tensor, height, width, self.vae_scale_factor)
+        tensor = (
+            tensor / self.vae.config.scaling_factor
+        ) + self.vae.config.shift_factor
+
+        with torch.no_grad():
+            image = self.vae.decode(tensor, return_dict=False)[0]
+
+        image = self.image_processor.postprocess(image, output_type="pil")
+
+        return image[0]

requirements.txt

@@ -0,0 +1,3 @@
+huggingface_hub
+diffusers
+