fix: pydantic models for http server working now. Fixes #380

.gitignore

@@ -30,3 +30,4 @@ tests/vastai_cli.py
 **/.eggs
 /img_size_memory_usage.csv
 /tests/test_cluster_output/
+/.env

README.md

@@ -493,6 +493,10 @@ A: The AI models are cached in `~/.cache/` (or `HUGGINGFACE_HUB_CACHE`). To dele
 
 ## ChangeLog
 
+**13.2.1**
+- fix: pydantic models for http server working now. Fixes #380
+- fix: install triton so annoying message is gone
+
 **13.2.0**
 - fix: allow tile_mode to be set to True or False for backward compatibility
 - fix: various pydantic issues have been resolved

imaginairy/http/stablestudio/models.py

@@ -119,5 +119,5 @@ class StableStudioBatchResponse(BaseModel):
     images: List[StableStudioImage]
 
 
-StableStudioInput.update_forward_refs()
+StableStudioInput.model_rebuild()
-StableStudioImage.update_forward_refs()
+StableStudioImage.model_rebuild()

imaginairy/http/utils.py

@@ -5,7 +5,7 @@ from imaginairy import imagine
 
 
 def generate_image(prompt):
-    """ImaginPrompt to generated image."""
+    """ImaginePrompt to generated image."""
     result = next(imagine([prompt]))
     img = result.images["generated"]
     img_io = BytesIO()
@@ -27,7 +27,7 @@ class Base64Bytes(bytes):
         yield cls.validate
 
     @classmethod
-    def validate(cls, v):
+    def validate(cls, v, info):
         if isinstance(v, bytes):
             return v
         if isinstance(v, str):

imaginairy/schema.py

@@ -18,7 +18,6 @@ from pydantic import (
     model_validator,
 )
 from pydantic_core import core_schema
-from pydantic_core.core_schema import FieldValidationInfo
 
 from imaginairy import config
 
@@ -90,6 +89,12 @@ class LazyLoadingImage:
         self._load_img()
         return getattr(self._img, key)
 
+    def __setstate__(self, state):
+        self.__dict__.update(state)
+
+    def __getstate__(self):
+        return self.__dict__
+
     def _load_img(self):
         if self._img is None:
             from PIL import Image, ImageOps
@@ -192,16 +197,10 @@ class LazyLoadingImage:
 
         shows filepath or url if available.
         """
+        try:
             return f"<LazyLoadingImage filepath={self._lazy_filepath} url={self._lazy_url}>"
-
+        except Exception as e:  # noqa
-
+            return f"<LazyLoadingImage RENDER EXCEPTION*{e}*>"
-#
-# LazyLoadingImage = Annotated[
-#     _LazyLoadingImage,
-#     AfterValidator(_LazyLoadingImage.validate),
-#     PlainSerializer(lambda i: str(i), return_type=str),
-#     WithJsonSchema({"type": "string"}, mode="serialization"),
-# ]
 
 
 class ControlNetInput(BaseModel):
@@ -218,7 +217,7 @@ class ControlNetInput(BaseModel):
     #     return v
 
     @field_validator("image_raw")
-    def image_raw_validate(cls, v, info: FieldValidationInfo):
+    def image_raw_validate(cls, v, info: core_schema.FieldValidationInfo):
         if info.data.get("image") is not None and v is not None:
             raise ValueError("You cannot specify both image and image_raw")
 
@@ -245,7 +244,7 @@ class WeightedPrompt(BaseModel):
         return f"{self.weight}*({self.text})"
 
 
-class ImaginePrompt(BaseModel):
+class ImaginePrompt(BaseModel, protected_namespaces=()):
     prompt: Optional[List[WeightedPrompt]] = Field(default=None, validate_default=True)
     negative_prompt: Optional[List[WeightedPrompt]] = Field(
         default=None, validate_default=True
@@ -403,7 +402,7 @@ class ImaginePrompt(BaseModel):
         return v
 
     @field_validator("control_inputs", mode="after")
-    def set_image_from_init_image(cls, v, info: FieldValidationInfo):
+    def set_image_from_init_image(cls, v, info: core_schema.FieldValidationInfo):
         v = v or []
         for control_input in v:
             if control_input.image is None and control_input.image_raw is None:
@@ -411,13 +410,13 @@ class ImaginePrompt(BaseModel):
         return v
 
     @field_validator("mask_image")
-    def validate_mask_image(cls, v, info: FieldValidationInfo):
+    def validate_mask_image(cls, v, info: core_schema.FieldValidationInfo):
         if v is not None and info.data.get("mask_prompt") is not None:
             raise ValueError("You can only set one of `mask_image` and `mask_prompt`")
         return v
 
     @field_validator("mask_prompt", "mask_image", mode="before")
-    def validate_mask_prompt(cls, v, info: FieldValidationInfo):
+    def validate_mask_prompt(cls, v, info: core_schema.FieldValidationInfo):
         if info.data.get("init_image") is None and v:
             raise ValueError("You must set `init_image` if you want to use a mask")
         return v
@@ -441,7 +440,7 @@ class ImaginePrompt(BaseModel):
         return v
 
     @field_validator("sampler_type", mode="after")
-    def validate_sampler_type(cls, v, info: FieldValidationInfo):
+    def validate_sampler_type(cls, v, info: core_schema.FieldValidationInfo):
         from imaginairy.samplers import SamplerName
 
         if v is None:
@@ -462,7 +461,7 @@ class ImaginePrompt(BaseModel):
         return v
 
     @field_validator("steps")
-    def validate_steps(cls, v, info: FieldValidationInfo):
+    def validate_steps(cls, v, info: core_schema.FieldValidationInfo):
         from imaginairy.samplers import SAMPLER_LOOKUP
 
         if v is None:
@@ -484,7 +483,7 @@ class ImaginePrompt(BaseModel):
         return self
 
     @field_validator("height", "width")
-    def validate_image_size(cls, v, info: FieldValidationInfo):
+    def validate_image_size(cls, v, info: core_schema.FieldValidationInfo):
         from imaginairy.model_manager import get_model_default_image_size
 
         if v is None:

requirements-dev.in

@@ -10,3 +10,5 @@ pytest-randomly
 pytest-sugar
 responses
 wheel
+
+-c tests/constraints.txt

requirements-dev.txt

@@ -16,7 +16,7 @@ anyio==3.7.1
     # via
     #   fastapi
     #   starlette
-astroid==2.15.6
+astroid==2.15.8
     # via pylint
 async-timeout==4.0.3
     # via aiohttp
@@ -42,13 +42,13 @@ click-help-colors==0.9.2
     # via imaginAIry (setup.py)
 click-shell==2.1
     # via imaginAIry (setup.py)
-contourpy==1.1.0
+contourpy==1.1.1
     # via matplotlib
 coverage==7.3.1
     # via -r requirements-dev.in
-cycler==0.11.0
+cycler==0.12.0
     # via matplotlib
-diffusers==0.20.2
+diffusers==0.21.3
     # via imaginAIry (setup.py)
 dill==0.3.7
     # via pylint
@@ -62,9 +62,9 @@ facexlib==0.3.0
     # via imaginAIry (setup.py)
 fairscale==0.4.13
     # via imaginAIry (setup.py)
-fastapi==0.103.1
+fastapi==0.103.2
     # via imaginAIry (setup.py)
-filelock==3.12.3
+filelock==3.12.4
     # via
     #   diffusers
     #   huggingface-hub
@@ -77,7 +77,7 @@ frozenlist==1.4.0
     # via
     #   aiohttp
     #   aiosignal
-fsspec[http]==2023.9.0
+fsspec[http]==2023.9.2
     # via
     #   huggingface-hub
     #   pytorch-lightning
@@ -87,7 +87,7 @@ ftfy==6.1.1
     #   open-clip-torch
 h11==0.14.0
     # via uvicorn
-huggingface-hub==0.17.1
+huggingface-hub==0.17.3
     # via
     #   diffusers
     #   open-clip-torch
@@ -98,7 +98,7 @@ idna==3.4
     #   anyio
     #   requests
     #   yarl
-imageio==2.31.3
+imageio==2.31.4
     # via imaginAIry (setup.py)
 importlib-metadata==6.8.0
     # via diffusers
@@ -120,10 +120,12 @@ lightning-utilities==0.9.0
     # via
     #   pytorch-lightning
     #   torchmetrics
-llvmlite==0.40.1
+llvmlite==0.41.0
     # via numba
 matplotlib==3.7.3
-    # via filterpy
+    # via
+    #   -c tests/constraints.txt
+    #   filterpy
 mccabe==0.7.0
     # via
     #   pylama
@@ -136,10 +138,11 @@ mypy-extensions==1.0.0
     # via
     #   black
     #   typing-inspect
-numba==0.57.1
+numba==0.58.0
     # via facexlib
 numpy==1.24.4
     # via
+    #   -c tests/constraints.txt
     #   contourpy
     #   diffusers
     #   facexlib
@@ -159,7 +162,7 @@ omegaconf==2.3.0
     # via imaginAIry (setup.py)
 open-clip-torch==2.20.0
     # via imaginAIry (setup.py)
-opencv-python==4.8.0.76
+opencv-python==4.8.1.78
     # via
     #   facexlib
     #   imaginAIry (setup.py)
@@ -178,7 +181,7 @@ pathspec==0.11.2
     # via
     #   black
     #   pycln
-pillow==10.0.0
+pillow==10.0.1
     # via
     #   diffusers
     #   facexlib
@@ -202,11 +205,11 @@ pycln==2.2.2
     # via -r requirements-dev.in
 pycodestyle==2.11.0
     # via pylama
-pydantic==2.3.0
+pydantic==2.4.2
     # via
     #   fastapi
     #   imaginAIry (setup.py)
-pydantic-core==2.6.3
+pydantic-core==2.10.1
     # via pydantic
 pydocstyle==6.3.0
     # via pylama
@@ -214,7 +217,7 @@ pyflakes==3.1.0
     # via pylama
 pylama==8.4.1
     # via -r requirements-dev.in
-pylint==2.17.5
+pylint==2.17.6
     # via -r requirements-dev.in
 pyparsing==3.1.1
     # via matplotlib
@@ -257,7 +260,7 @@ requests==2.31.0
     #   transformers
 responses==0.23.3
     # via -r requirements-dev.in
-ruff==0.0.288
+ruff==0.0.291
     # via -r requirements-dev.in
 safetensors==0.3.3
     # via
@@ -312,7 +315,7 @@ torch==1.13.1
     #   torchvision
 torchdiffeq==0.2.3
     # via imaginAIry (setup.py)
-torchmetrics==1.1.2
+torchmetrics==1.2.0
     # via
     #   imaginAIry (setup.py)
     #   pytorch-lightning
@@ -330,18 +333,17 @@ tqdm==4.66.1
     #   open-clip-torch
     #   pytorch-lightning
     #   transformers
-transformers==4.33.1
+transformers==4.33.3
     # via imaginAIry (setup.py)
 typer==0.9.0
     # via pycln
-types-pyyaml==6.0.12.11
+types-pyyaml==6.0.12.12
     # via responses
-typing-extensions==4.7.1
+typing-extensions==4.8.0
     # via
     #   astroid
     #   black
     #   fastapi
-    #   filelock
     #   huggingface-hub
     #   libcst
     #   lightning-utilities
@@ -355,13 +357,13 @@ typing-extensions==4.7.1
     #   uvicorn
 typing-inspect==0.9.0
     # via libcst
-urllib3==2.0.4
+urllib3==2.0.5
     # via
     #   requests
     #   responses
 uvicorn==0.23.2
     # via imaginAIry (setup.py)
-wcwidth==0.2.6
+wcwidth==0.2.7
     # via ftfy
 wheel==0.41.2
     # via -r requirements-dev.in
@@ -369,5 +371,5 @@ wrapt==1.15.0
     # via astroid
 yarl==1.9.2
     # via aiohttp
-zipp==3.16.2
+zipp==3.17.0
     # via importlib-metadata

setup.py

@@ -103,9 +103,10 @@ setup(
         "scipy<1.11",
         "timm>=0.4.12,!=0.9.0,!=0.9.1",  # for vendored blip
         "torchdiffeq>=0.2.0",
-        "transformers>=4.19.2",
         "torchmetrics>=0.6.0",
         "torchvision>=0.13.1",
+        "transformers>=4.19.2",
+        "triton>=2.0.0; sys_platform!='darwin' and platform_machine!='aarch64'",
         "kornia>=0.6",
         "uvicorn>=0.16.0",
         "xformers>=0.0.16; sys_platform!='darwin' and platform_machine!='aarch64'",

tests/constraints.txt

@@ -0,0 +1,3 @@
+# held back for python 3.8 compatability
+matplotlib<3.8.0
+numpy<1.25.0

tests/test_schema/test_lazy_load_image.py

@@ -101,3 +101,14 @@ def test_image_deserialization(red_path, red_url):
     for row in rows:
         obj = TestModel.model_validate(row)
         assert obj.header_img.size == (512, 512)
+
+
+def test_image_state(red_path):
+    """I dont remember what this fixes. Maybe the ability of pydantic to copy an object?."""
+    img = LazyLoadingImage(filepath=red_path)
+
+    # bypass init
+    img2 = LazyLoadingImage.__new__(LazyLoadingImage)
+    img2.__setstate__(img.__getstate__())
+
+    assert repr(img) == repr(img2)