ai-simple-engine-diffusion 0.0.6__tar.gz → 0.0.8__tar.gz

{ai_simple_engine_diffusion-0.0.6 → ai_simple_engine_diffusion-0.0.8}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: ai-simple-engine-diffusion
-Version: 0.0.6
+Version: 0.0.8
 Summary: AI Simple Engine Diffusion Module
 License-File: LICENSE
 Author: danialcala94

{ai_simple_engine_diffusion-0.0.6 → ai_simple_engine_diffusion-0.0.8}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "ai-simple-engine-diffusion"
-version = "0.0.6"
+version = "0.0.8"
 description = "AI Simple Engine Diffusion Module"
 authors = [
     {name = "danialcala94", email = "danielalcalavalera@gmail.com"}

{ai_simple_engine_diffusion-0.0.6 → ai_simple_engine_diffusion-0.0.8}/src/ai_simple_engine_diffusion/model/info/diffusion_model_info.py

@@ -1,5 +1,6 @@
 from ai_simple_engine_diffusion.scheduler.spec.base import SchedulerSpec
 from ai_simple_engine_diffusion.types.latent_space import LatentSpace
+from ai_simple_engine_diffusion.model.info.tensor_output_spec import TensorOutputSpec
 from ai_simple_engine.models.info.abstract import ModelInfo
 from dataclasses import dataclass
 # from typing import Union
@@ -23,6 +24,10 @@ class DiffusionModelInfo(
     The scheduler that must be used for this
     specific model: 'euler', 'flow_match', etc.
     """
+    tensor_output_spec: TensorOutputSpec
+    """
+    The specs about the image we get as output.
+    """
 
     def latent_space(
         self,

ai_simple_engine_diffusion-0.0.8/src/ai_simple_engine_diffusion/model/info/tensor_output_spec.py

@@ -0,0 +1,19 @@
+from dataclasses import dataclass
+
+
+@dataclass(frozen = True)
+class TensorOutputSpec:
+    """
+    Specifications about the output tensor that
+    a model gives, which will let us transform
+    into a 'normalized' tensor.
+    """
+
+    value_range: tuple[float, float]
+    channels: int = 3
+    color_space: str = 'rgb'
+    # Probably soon:
+    # layout (NCHW / NHWC)
+    # dtype (float16, float32, uint8)
+    # rango ([-1,1], [0,1], [0,255])
+    # espacio de color (RGB, BGR, YCbCr)

ai_simple_engine_diffusion-0.0.8/src/ai_simple_engine_diffusion/types/latent_space.py

@@ -0,0 +1,43 @@
+from dataclasses import dataclass
+
+
+@dataclass(frozen = True)
+class LatentSpace:
+    """
+    The space in which all the internal vectors
+    (embeddings) live, where they are plotted and
+    compared. The internal geometry created by the
+    model's parameters.
+
+    This is the CANVAS (space) in which we plot the
+    embeddings, and is only affected when tunning
+    parameters or training the model.
+    """
+
+    channels: int
+    width: int
+    height: int
+    batch_size: int = 1
+
+    def to_shape(
+        self
+    ) -> tuple[int, int, int, int]:
+        """
+        Transform the `LatentSpace` instance into a
+        shape, useful to be sent to the noise
+        generator. The shape is a tuple like this:
+        ```
+        (
+            self.batch_size,
+            self.channels,
+            self.height,
+            self.width
+        )
+        ```
+        """
+        return (
+            self.batch_size,
+            self.channels,
+            self.height,
+            self.width
+        )

ai_simple_engine_diffusion-0.0.6/src/ai_simple_engine_diffusion/types/latent_space.py

@@ -1,20 +0,0 @@
-from dataclasses import dataclass
-
-
-@dataclass(frozen = True)
-class LatentSpace:
-    """
-    The space in which all the internal vectors
-    (embeddings) live, where they are plotted and
-    compared. The internal geometry created by the
-    model's parameters.
-
-    This is the CANVAS (space) in which we plot the
-    embeddings, and is only affected when tunning
-    parameters or training the model.
-    """
-
-    channels: int
-    width: int
-    height: int
-    batch_size: int = 1