leap-model-parser 0.1.242__tar.gz → 0.1.243.dev1__tar.gz

{leap_model_parser-0.1.242 → leap_model_parser-0.1.243.dev1}/PKG-INFO

@@ -1,7 +1,8 @@
-Metadata-Version: 2.3
+Metadata-Version: 2.1
 Name: leap-model-parser
-Version: 0.1.242
+Version: 0.1.243.dev1
 Summary: 
+Home-page: https://github.com/tensorleap/leap-model-parser
 License: MIT
 Author: idan
 Author-email: idan.yogev@tensorleap.ai
@@ -20,7 +21,6 @@ Requires-Dist: onnx2kerastl (==0.0.186)
 Requires-Dist: tensorflow (==2.12.0) ; platform_machine == "x86_64"
 Requires-Dist: tensorflow-io-gcs-filesystem (==0.34.0)
 Requires-Dist: tensorflow-macos (==2.12.0) ; platform_machine == "arm64"
-Project-URL: Homepage, https://github.com/tensorleap/leap-model-parser
 Project-URL: Repository, https://github.com/tensorleap/leap-model-parser
 Description-Content-Type: text/markdown
 

{leap_model_parser-0.1.242 → leap_model_parser-0.1.243.dev1}/leap_model_parser/model_parser.py

@@ -34,7 +34,46 @@ if spec is not None:
     import onnx  # type: ignore
 
     onnx_imported = True
-
+#Deal with broken models
+GIT_LFS_CONTENT = b"version https://git-lfs.github.com/spec/v1"
+
+class InvalidModelFile(Exception):
+    pass
+
+def _is_git_lfs_pointer(path: str) -> bool:
+    try:
+        with open(path, "rb") as f:
+            return f.read(128).startswith(GIT_LFS_CONTENT)
+    except OSError:
+        return False
+
+
+def _text_snippet(path: str, n: int = 200) -> Optional[str]:
+    try:
+        with open(path, "rb") as f:
+            b = f.read(n)
+        if b"\x00" in b:
+            return None
+        s = b.decode("utf-8", errors="replace").strip()
+        return s or None
+    except OSError:
+        return None
+
+
+def _raise_model_file_error(path: str, kind: str, original_exc: Exception) -> None:
+    suffix = Path(path).suffix
+    if _is_git_lfs_pointer(path):
+        raise InvalidModelFile(
+            f"This {suffix} file is a Git LFS pointer, not the actual model binary. "
+            "Run `git lfs pull` (or re-download the artifact) to fetch the real file."
+        ) from original_exc
+    snippet = _text_snippet(path)
+    if snippet is not None:
+        raise InvalidModelFile(
+            f"The provided model is not a real {kind} model binary although its suffix is {suffix}. "
+            f"File's content starts with:\n{snippet}."
+        ) from original_exc
+    raise InvalidModelFile(f"Failed to open {kind} model file: {original_exc}") from original_exc
 
 class ModelParser:
     def __init__(self, should_transform_inputs_and_outputs=False,
@@ -135,7 +174,10 @@ class ModelParser:
         if not onnx_imported:
             raise OnnxUnsupported()
 
-        onnx_model = onnx.load_model(file_path)
+        try:
+            onnx_model = onnx.load_model(file_path)
+        except Exception as e:
+            _raise_model_file_error(file_path, "onnx", e)
         input_all = [_input.name for _input in onnx_model.graph.input]
         input_initializer = [
             node.name for node in onnx_model.graph.initializer]
@@ -153,10 +195,9 @@ class ModelParser:
         try:
             return load_model(file_path, custom_objects=custom_objects, compile=False)
         except OSError as e:
-            if 'signature' in str(e):
-                raise Exception('Unable to open model file. The model might be corrupted or not a valid.')
-            else:
-                raise e
+            if "signature" in str(e):
+                _raise_model_file_error(file_path, "keras", e)
+            raise
 
     def convert_h5_model(self, file_path: str) -> Tuple[Dict[str, Node], Model, Optional[str]]:
         imported_model = self._load_keras_model_with_custom_layers(file_path)

{leap_model_parser-0.1.242 → leap_model_parser-0.1.243.dev1}/pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "leap-model-parser"
-version = "0.1.242"
+version = "0.1.243.dev1"
 description = ""
 authors = ["idan <idan.yogev@tensorleap.ai>"]
 license = "MIT"