Fix the documented examples and defaults that do not match the code (#1427)

Signed-off-by: Onuralp SEZER <thunderbirdtr@gmail.com>

Author: onuralpszr

CONTRIBUTING.md

@@ -26,7 +26,7 @@ source .venv/bin/activate  # On Windows: .venv\Scripts\activate
 
 ```bash
 # Install core + dev dependencies
-uv sync --extra dev
+uv sync --group dev
 
 # For testing specific models, install their dependencies.
 ```
@@ -92,7 +92,7 @@ If the CI build fails due to formatting:
 3. Install dev dependencies:
 
    ```bash
-   uv sync --extra dev
+   uv sync --group dev
    ```
 
 4. Fix formatting:

docs/cli.md

@@ -89,8 +89,8 @@ sahi predict --slice_width 512 --slice_height 512 \
 
 **Match Metrics:**
 
-- `--postprocess_match_metric IOS` - Intersection over smaller area
-- `--postprocess_match_metric IOU` - Intersection over union (default)
+- `--postprocess_match_metric IOS` - Intersection over smaller area (default)
+- `--postprocess_match_metric IOU` - Intersection over union
 
 **Additional Options:**
 
@@ -127,8 +127,7 @@ sahi predict --dataset_json_path dataset.json \
   --model_path path/to/model
 ```
 
-Predictions will be exported as a COCO JSON file to
-`runs/predict/exp/results.json`. You can then use:
+Adding `--dataset_json_path` also exports predictions as a COCO JSON file to `runs/predict/exp/result.json`. You can then use:
 
 - `sahi coco evaluate` - Calculate COCO evaluation metrics
 - `sahi coco analyse` - Generate detailed error analysis plots
@@ -367,7 +366,6 @@ Display your currently installed SAHI version.
 
 ```bash
 sahi version
-0.11.22
 ```
 
 ---

docs/guides/models.md

@@ -447,9 +447,9 @@ All models accept these parameters in `AutoDetectionModel.from_pretrained()`:
 | `model_type` | str | Framework name (see sections above) |
 | `model_path` | str | Path to weights file or model name |
 | `config_path` | str | Config file path (MMDetection, Detectron2) |
-| `confidence_threshold` | float | Minimum score to keep a detection (default: 0.25) |
+| `confidence_threshold` | float | Minimum score to keep a detection (default: 0.3) |
 | `device` | str | `"cpu"`, `"cuda:0"`, `"mps"`, etc. |
-| `category_mapping` | dict | Map category IDs to names: `{0: "car", 1: "person"}` |
+| `category_mapping` | dict | Map category IDs to names, keys are strings: `{"0": "car", "1": "person"}` |
 | `category_remapping` | dict | Remap category names after inference |
 | `image_size` | int | Override model input resolution |
 | `load_at_init` | bool | Load weights immediately (default: True) |

docs/notebooks.md

@@ -40,6 +40,6 @@ Clone the repository and run notebooks with Jupyter:
 ```bash
 git clone https://github.com/obss/sahi.git
 cd sahi
-pip install -e ".[dev]"
+uv sync --group dev
 jupyter notebook demo/
 ```

docs/postprocess/backends.md

@@ -24,10 +24,10 @@ right backend depends on your hardware and installed packages.
 
 By default SAHI automatically picks the best available backend at runtime:
 
-1. **torchvision** — if `torchvision` is installed _and_ a GPU is present
+1. **torchvision**: if `torchvision` is installed _and_ a GPU is present
    (CUDA, or Apple MPS on Apple Silicon).
-2. **numba** — if the `numba` package is installed.
-3. **numpy** — always available as the final fallback.
+2. **numba**: if the `numba` package is installed.
+3. **numpy**: always available as the final fallback.
 
 ```python
 from sahi.postprocess.backends import get_postprocess_backend
@@ -103,11 +103,11 @@ predictions = np.array([
     [300, 300, 400, 400, 0.90, 1],
 ])
 
-# Global NMS — all categories compete together
+# Global NMS, all categories compete together
 keep = nms(predictions, match_metric="IOU", match_threshold=0.5)
 print(predictions[keep])
 
-# Per-category NMS — class 0 and class 1 are treated independently
+# Per-category NMS, class 0 and class 1 are treated independently
 keep = batched_nms(predictions, match_metric="IOU", match_threshold=0.5)
 print(predictions[keep])
 ```
@@ -147,19 +147,19 @@ by `get_sliced_prediction` via the `postprocess_type` argument:
 ```python
 from sahi.postprocess.combine import NMSPostprocess, NMMPostprocess, GreedyNMMPostprocess
 
-# NMS — keep the best box, discard the rest
+# NMS, keep the best box, discard the rest
 postprocessor = NMSPostprocess(
     match_threshold=0.5,
     match_metric="IOU",
     class_agnostic=True,   # False → per-category
 )
 filtered = postprocessor(object_prediction_list)
 
-# Greedy NMM — merge overlapping boxes (fast)
+# Greedy NMM, merge overlapping boxes (fast)
 postprocessor = GreedyNMMPostprocess(match_threshold=0.5)
 merged = postprocessor(object_prediction_list)
 
-# Full NMM — transitive merging
+# Full NMM, transitive merging
 postprocessor = NMMPostprocess(match_threshold=0.5)
 merged = postprocessor(object_prediction_list)
 ```

docs/predict.md

@@ -131,11 +131,9 @@ for i, preds in enumerate(detection_model.object_prediction_list_per_image):
         print(pred.category.name, pred.score.value, pred.bbox.to_xyxy())
 ```
 
-!!! note "Single-image compatibility" The existing `object_prediction_list`
+!!! note "Single-image compatibility"
 
-property is unchanged and returns predictions for the first image, so code that
-uses `perform_inference` + `convert_original_predictions` +
-`object_prediction_list` continues to work without modification.
+    The existing `object_prediction_list` property is unchanged and returns predictions for the first image, so code that uses `perform_inference` + `convert_original_predictions` + `object_prediction_list` continues to work without modification.
 
 ## Progress-Bar
 
@@ -175,11 +173,11 @@ result = get_sliced_prediction(
 )
 ```
 
-!!! tip "Notes" - `progress_bar` and `progress_callback` can be used together.
+!!! tip "Notes"
 
-When both are provided, the tqdm bar will display and the callback will be
-called after each slice group is processed. - The `progress_callback` is called
-with 1-based indices (i.e. first call will be `(1, total)`).
+    `progress_bar` and `progress_callback` can be used together. When both are provided, the tqdm bar will display and the callback will be called after each slice group is processed.
+
+    The `progress_callback` is called with 1-based indices, so the first call will be `(1, total)`.
 
 ## Exclude custom classes on inference
 
@@ -230,12 +228,9 @@ result.export_visuals(
     export_dir="outputs/",
     text_size=1.0,  # Size of the class label text
     rect_th=2,      # Thickness of bounding box lines
-    text_th=2,      # Thickness of the text
     hide_labels=False,  # Set True to hide class labels
     hide_conf=False,    # Set True to hide confidence scores
-    color=(255, 0, 0),  # Custom color in RGB format (red in this example)
     file_name="custom_visualization",
-    export_format="jpg"  # Supports 'jpg' and 'png'
 )
 
 # Export as COCO format annotations