Add scratch of parsing for new config in inference-exp

Author: PawelPeczek-Roboflow

inference_experimental/inference_exp/models/common/roboflow/model_packages.py

@@ -2,11 +2,12 @@
 from collections import namedtuple
 from dataclasses import dataclass
 from enum import Enum
-from typing import List, Optional
+from typing import Annotated, List, Literal, Optional, Tuple, Union
 
 from inference_exp.entities import ImageDimensions
 from inference_exp.errors import CorruptedModelPackageError
 from inference_exp.utils.file_system import read_json, stream_file_lines
+from pydantic import BaseModel, Field
 
 
 def parse_class_names_file(class_names_path: str) -> List[str]:
@@ -259,3 +260,178 @@ def parse_class_map_from_environment_file(environment_file_path: str) -> List[st
             f"verify its consistency in docs.",
             help_url="https://todo",
         ) from error
+
+
+# non_nms_config = {
+#     "image_pre_processing": image_pre_processing,
+#     "network_input": {
+#         "training_input_size": {"height": self.imgsz, "width": self.imgsz},
+#         "dynamic_spatial_size_supported": False,
+#         "color_mode": "rgb",
+#         "resize_mode": resize_mode,
+#         "padding_value": padding_value,
+#         "input_channels": 3,
+#         "scaling_factor": 255,
+#         "normalization": None,
+#     },
+#     "post_processing": {
+#         "type": "nms",
+#         "fused": False,
+#     }
+# }
+
+
+class AutoOrient(BaseModel):
+    enabled: bool
+
+
+class StaticCrop(BaseModel):
+    enabled: bool
+    x_min: int
+    x_max: int
+    y_min: int
+    y_max: int
+
+
+class ContrastType(Enum, str):
+    ADAPTIVE_EQUALIZATION = "Adaptive Equalization"
+    CONTRAST_STRETCHING = "Contrast Stretching"
+    HISTOGRAM_EQUALIZATION = "Histogram Equalization"
+
+
+class Contrast(BaseModel):
+    enabled: bool
+    type: ContrastType
+
+
+class Grayscale(BaseModel):
+    enabled: bool
+
+
+class ImagePreProcessing(BaseModel):
+    auto_orient: Optional[AutoOrient] = Field(alias="auto-orient", default=None)
+    static_crop: Optional[StaticCrop] = Field(alias="static-crop", default=None)
+    contrast: Optional[Contrast] = Field(default=None)
+    grayscale: Optional[Grayscale] = Field(default=None)
+
+
+class TrainingInputSize(BaseModel):
+    height: int
+    width: int
+
+
+class DivisiblePadding(BaseModel):
+    type: Literal["pad-to-be-divisible"]
+    value: int
+
+
+class AnySizePadding(BaseModel):
+    type: Literal["any-size"]
+
+
+class ColorMode(Enum, str):
+    BGR = "bgr"
+    RGB = "rgb"
+
+
+# non_nms_config = {
+#     "image_pre_processing": image_pre_processing,
+#     "network_input": {
+#         "training_input_size": {"height": self.imgsz, "width": self.imgsz},
+#         "dynamic_spatial_size_supported": False,
+#         "color_mode": "rgb",
+#         "resize_mode": resize_mode,
+#         "padding_value": padding_value,
+#         "input_channels": 3,
+#         "scaling_factor": 255,
+#         "normalization": None,
+#     },
+#     "post_processing": {
+#         "type": "nms",
+#         "fused": False,
+#     }
+# }
+#
+#  "post_processing": {
+#                 "type": "nms",
+#                 "fused": True,
+#                 "nms_parameters": nms_parameters_cleaned,
+#             }
+class ResizeMode(Enum, str):
+    STRETCH_TO = "Stretch to"
+    LETTERBOX = "letterbox"
+    CENTER_CROP = "center-crop"
+    FIT_LONGER_EDGE = "fit-longer-edge"
+    LETTERBOX_REFLECT_EDGES = "letterbox-reflect-edges"
+
+
+Number = Union[int, float]
+
+
+class NetworkInput(BaseModel):
+    training_input_size: TrainingInputSize
+    dynamic_spatial_size_supported: bool
+    dynamic_spatial_size_mode: Optional[Union[DivisiblePadding, AnySizePadding]] = (
+        Field(discriminator="type", default=None)
+    )
+    color_mode: ColorMode
+    resize_mode: ResizeMode
+    padding_value: Optional[int]
+    input_channels: int
+    scaling_factor: Optional[Number] = Field(default=None)
+    normalization: Optional[
+        Tuple[Tuple[Number, Number, Number], Tuple[Number, Number, Number]]
+    ] = Field(default=None)
+
+
+class ForwardPassConfiguration(BaseModel):
+    max_dynamic_batch_size: int
+
+
+class FusedNMSParameters(BaseModel):
+    max_detections: int
+    confidence_threshold: float
+    iou_threshold: float
+    class_agnostic: int
+
+
+class NMSPostProcessing(BaseModel):
+    type: Literal["nms"]
+    fused: bool
+    nms_parameters: Optional[FusedNMSParameters] = Field(default=None)
+
+
+class SigmoidPostProcessing(BaseModel):
+    type: Literal["sigmoid"]
+    fused: bool
+
+
+class SoftMaxPostProcessing(BaseModel):
+    type: Literal["softmax"]
+    fused: bool
+
+
+class InferenceConfig(BaseModel):
+    image_pre_processing: ImagePreProcessing
+    network_input: NetworkInput
+    forward_pass: Optional[ForwardPassConfiguration] = Field(default=None)
+    post_processing: Optional[
+        Union[NMSPostProcessing, SoftMaxPostProcessing, SigmoidPostProcessing]
+    ] = Field(default=None, discriminator="type")
+
+
+def parse_inference_config(config_path: str) -> None:
+    try:
+        parsed_config = read_json(path=config_path)
+        if not isinstance(parsed_config, dict):
+            raise ValueError(
+                f"Expected config format is dict, found {type(parsed_config)} instead"
+            )
+    except (IOError, OSError, ValueError) as error:
+        raise CorruptedModelPackageError(
+            message=f"Inference config file of the model package is malformed: "
+            f"{error}. In case that the package is "
+            f"hosted on the Roboflow platform - contact support. If you created model package manually, please "
+            f"verify its consistency in docs.",
+            help_url="https://todo",
+        ) from error

inference_experimental/inference_exp/models/florence2/florence2_hf.py

@@ -1,17 +1,17 @@
-from typing import List, Literal, Optional, Tuple, Union
 import os
+from typing import List, Literal, Optional, Tuple, Union
 
 import cv2
 import numpy as np
 import torch
-from peft import LoraConfig, PeftModel
 from inference_exp import Detections, InstanceDetections
 from inference_exp.configuration import DEFAULT_DEVICE
 from inference_exp.entities import ImageDimensions
 from inference_exp.errors import ModelRuntimeError
 from inference_exp.models.common.roboflow.pre_processing import (
     extract_input_images_dimensions,
 )
+from peft import LoraConfig, PeftModel
 from transformers import AutoModelForCausalLM, AutoProcessor
 
 GRANULARITY_2TASK = {

inference_experimental/inference_exp/models/yolov8/yolov8_object_detection_onnx.py

@@ -78,7 +78,7 @@ def from_pretrained(
             model_package_dir=model_name_or_path,
             elements=[
                 "class_names.txt",
-                "environment.json",
+                "inference_config.json",
                 "weights.onnx",
             ],
         )

inference_experimental/tests/integration_tests/models/test_florence2_predictions.py

@@ -1,7 +1,6 @@
 import numpy as np
 import pytest
 import torch
-
 from inference_exp.models.florence2.florence2_hf import Florence2HF
 
 

inference_experimental/tests/integration_tests/models/test_florence2_preprocessing.py

@@ -1,7 +1,6 @@
 import numpy as np
 import pytest
 import torch
-
 from inference_exp.models.florence2.florence2_hf import Florence2HF