Merge branch 'main' into dependabot/pip/idna-3.7

Author: capjamesg

.devcontainer/devcontainer.json

@@ -8,6 +8,7 @@
         "vscode": {
             "extensions": [
                 "ms-python.python",
+                "ms-python.debugpy",
                 "ms-python.black-formatter"
             ],
             "settings": {

CONTRIBUTING.md

@@ -5,7 +5,13 @@ If you want to extend our Python library or if you find a bug, please open a PR!
 
 Also be sure to test your code the `unittest` command at the `/root` level directory.
 
-Run tests:
+### Devcontainer
+
+This project comes with a [convenient devcontainer](https://www.loom.com/share/a183c4a351ed4700a79476fedf08ab9b) that makes it easier to run tests and has black configured to run on save.
+
+On rare occasions a full rebuild is needed, you can do it in VSCode by pressing `Ctrl+Shift+P` and running `Dev Containers: Rebuild Container`.
+
+### Tests
 
 ```bash
 python -m unittest
@@ -39,4 +45,12 @@ make check_code_quality
 
 **Note** These tests will be run automatically when you commit thanks to git hooks.
 
-**Note** This project also comes with a [convenient devcontainer](https://www.loom.com/share/a183c4a351ed4700a79476fedf08ab9b) that makes it easier to run tests and has black configured to run on save.
+### Docs
+
+The docs can be built with `mkdocs serve`.
+
+Before that, install the dependencies:
+
+```python
+python -m pip install mkdocs mkdocs-material mkdocstrings mkdocstrings[python]
+```

requirements.txt

@@ -1,11 +1,11 @@
-certifi==2023.7.22
+certifi
 chardet==4.0.0
-cycler==0.10.0
 idna==3.7
+cycler
 kiwisolver>=1.3.1
 matplotlib
 numpy>=1.18.5
-opencv-python-headless==4.8.0.74
+opencv-python-headless==4.10.0.84
 Pillow>=7.1.2
 python-dateutil
 python-dotenv

roboflow/__init__.py

@@ -14,7 +14,7 @@
 from roboflow.models import CLIPModel, GazeModel  # noqa: F401
 from roboflow.util.general import write_line
 
-__version__ = "1.1.27"
+__version__ = "1.1.32"
 
 
 def check_key(api_key, model, notebook, num_retries=0):

roboflow/config.py

@@ -75,6 +75,7 @@ def get_conditional_configuration_variable(key, default):
 DEFAULT_JOB_NAME = "Annotated via API"
 
 RF_WORKSPACES = get_conditional_configuration_variable("workspaces", default={})
+TQDM_DISABLE = os.getenv("TQDM_DISABLE", None)
 
 
 def load_roboflow_api_key(workspace_url=None):

roboflow/core/project.py

@@ -528,9 +528,9 @@ def single_upload(
 
     def _annotation_params(self, annotation_path):
         annotation_name, annotation_string = None, None
-        if isinstance(annotation_path, dict):
+        if isinstance(annotation_path, dict) and annotation_path.get("rawText"):
             annotation_name = annotation_path["name"]
-            annotation_string = json.dumps(annotation_path["parsed"])
+            annotation_string = annotation_path["rawText"]
         elif os.path.exists(annotation_path):
             with open(annotation_path, "r"):
                 annotation_string = open(annotation_path, "r").read()

roboflow/core/version.py

@@ -17,6 +17,7 @@
     API_URL,
     APP_URL,
     DEMO_KEYS,
+    TQDM_DISABLE,
     TYPE_CLASSICATION,
     TYPE_INSTANCE_SEGMENTATION,
     TYPE_KEYPOINT_DETECTION,
@@ -432,11 +433,15 @@ def deploy(self, model_type: str, model_path: str, filename: str = "weights/best
             filename (str, optional): The name of the weights file. Defaults to "weights/best.pt".
         """
 
-        supported_models = ["yolov5", "yolov7-seg", "yolov8", "yolov9", "yolonas"]
+        supported_models = ["yolov5", "yolov7-seg", "yolov8", "yolov9", "yolonas", "paligemma", "yolov10"]
 
         if not any(supported_model in model_type for supported_model in supported_models):
             raise (ValueError(f"Model type {model_type} not supported. Supported models are" f" {supported_models}"))
 
+        if "paligemma" in model_type:
+            self.deploy_paligemma(model_type, model_path, filename)
+            return
+
         if "yolonas" in model_type:
             self.deploy_yolonas(model_type, model_path, filename)
             return
@@ -454,6 +459,17 @@ def deploy(self, model_type: str, model_path: str, filename: str = "weights/best
 
             print_warn_for_wrong_dependencies_versions([("ultralytics", "==", "8.0.196")], ask_to_continue=True)
 
+        elif "yolov10" in model_type:
+            try:
+                import torch
+                import ultralytics
+
+            except ImportError:
+                raise (
+                    "The ultralytics python package is required to deploy yolov10"
+                    " models. Please install it with `pip install ultralytics`"
+                )
+
         elif "yolov5" in model_type or "yolov7" in model_type or "yolov9" in model_type:
             try:
                 import torch
@@ -474,7 +490,7 @@ def deploy(self, model_type: str, model_path: str, filename: str = "weights/best
             class_names.sort(key=lambda x: x[0])
             class_names = [x[1] for x in class_names]
 
-        if "yolov8" in model_type:
+        if "yolov8" in model_type or "yolov10" in model_type:
             # try except for backwards compatibility with older versions of ultralytics
             if "-cls" in model_type:
                 nc = model["model"].yaml["nc"]
@@ -548,6 +564,57 @@ def deploy(self, model_type: str, model_path: str, filename: str = "weights/best
 
         self.upload_zip(model_type, model_path)
 
+    def deploy_paligemma(
+        self, model_type: str, model_path: str, filename: str = "fine-tuned-paligemma-3b-pt-224.f16.npz"
+    ) -> None:
+        # Check if model_path exists
+        if not os.path.exists(model_path):
+            raise FileNotFoundError(f"Model path {model_path} does not exist.")
+        model_files = os.listdir(model_path)
+        print(f"Model files found in {model_path}: {model_files}")
+
+        files_to_deploy = []
+
+        # Find first .npz file in model_path
+        npz_filename = next((file for file in model_files if file.endswith(".npz")), None)
+        if any([file.endswith(".safetensors") for file in model_files]):
+            print("Found .safetensors file in model path. Deploying PyTorch PaliGemma model.")
+            necessary_files = [
+                "config.json",
+                "generation_config.json",
+                "model.safetensors.index.json",
+                "preprocessor_config.json",
+                "special_tokens_map.json",
+                "tokenizer_config.json",
+                "tokenizer.json",
+            ]
+            for file in necessary_files:
+                if file not in model_files:
+                    print("Missing necessary file", file)
+                    res = input("Do you want to continue? (y/n)")
+                    if res.lower() != "y":
+                        exit(1)
+            for file in model_files:
+                files_to_deploy.append(file)
+        elif npz_filename is not None:
+            print(f"Found .npz file {npz_filename} in model path. Deploying JAX PaliGemma model.")
+            files_to_deploy.append(npz_filename)
+        else:
+            raise FileNotFoundError(f"No .npz or .safetensors file found in model path {model_path}.")
+
+        if len(files_to_deploy) == 0:
+            raise FileNotFoundError(f"No valid files found in model path {model_path}.")
+        print(f"Zipping files for deploy: {files_to_deploy}")
+
+        import tarfile
+
+        with tarfile.open(os.path.join(model_path, "roboflow_deploy.tar"), "w") as tar:
+            for file in files_to_deploy:
+                tar.add(os.path.join(model_path, file), arcname=file)
+
+        print("Uploading to Roboflow... May take several minutes.")
+        self.upload_zip(model_type, model_path, "roboflow_deploy.tar")
+
     def deploy_yolonas(self, model_type: str, model_path: str, filename: str = "weights/best.pt") -> None:
         try:
             import torch
@@ -613,7 +680,7 @@ def deploy_yolonas(self, model_type: str, model_path: str, filename: str = "weig
 
         self.upload_zip(model_type, model_path)
 
-    def upload_zip(self, model_type: str, model_path: str):
+    def upload_zip(self, model_type: str, model_path: str, model_file_name: str = "roboflow_deploy.zip"):
         res = requests.get(
             f"{API_URL}/{self.workspace}/{self.project}/{self.version}"
             f"/uploadModel?api_key={self.__api_key}&modelType={model_type}&nocache=true"
@@ -632,7 +699,7 @@ def upload_zip(self, model_type: str, model_path: str):
 
         res = requests.put(
             res.json()["url"],
-            data=open(os.path.join(model_path, "roboflow_deploy.zip"), "rb"),
+            data=open(os.path.join(model_path, model_file_name), "rb"),
         )
         try:
             res.raise_for_status()
@@ -685,9 +752,10 @@ def bar_progress(current, total, width=80):
             # write the zip file to the desired location
             with open(location + "/roboflow.zip", "wb") as f:
                 total_length = int(response.headers.get("content-length"))
+                desc = None if TQDM_DISABLE else f"Downloading Dataset Version Zip in {location} to {format}:"
                 for chunk in tqdm(
                     response.iter_content(chunk_size=1024),
-                    desc=f"Downloading Dataset Version Zip in {location} to {format}:",
+                    desc=desc,
                     total=int(total_length / 1024) + 1,
                 ):
                     if chunk:
@@ -711,10 +779,11 @@ def __extract_zip(self, location, format):
         Raises:
             RuntimeError: If there is an error unzipping the file
         """  # noqa: E501 // docs
+        desc = None if TQDM_DISABLE else f"Extracting Dataset Version Zip to {location} in {format}:"
         with zipfile.ZipFile(location + "/roboflow.zip", "r") as zip_ref:
             for member in tqdm(
                 zip_ref.infolist(),
-                desc=f"Extracting Dataset Version Zip to {location} in {format}:",
+                desc=desc,
             ):
                 try:
                     zip_ref.extract(member, location)

roboflow/core/workspace.py

@@ -341,8 +341,8 @@ def _upload_image(imagedesc):
             labelmap = None
             annotationdesc = imagedesc.get("annotationfile")
             if annotationdesc:
-                if annotationdesc.get("parsed"):
-                    annotation_path = {"name": annotationdesc["name"], "parsed": annotationdesc["parsed"]}
+                if annotationdesc.get("rawText"):
+                    annotation_path = annotationdesc
                 else:
                     annotation_path = f"{location}{annotationdesc['file']}"
                 labelmap = annotationdesc.get("labelmap")