[Releases 2.13] Add support for legacy OpenAI CLIP model (#1107)

marqo-ai · Jan 31, 2025 · c05a794 · c05a794
1 parent 8980c62
commit c05a794
Show file tree

Hide file tree

Showing 2 changed files with 14 additions and 2 deletions.
diff --git a/src/marqo/core/inference/embedding_models/open_clip_model.py b/src/marqo/core/inference/embedding_models/open_clip_model.py
@@ -210,7 +210,8 @@ def _load_model_and_image_preprocessor_from_open_clip_repo(self) -> Tuple[torch.
 
     def _load_tokenizer_from_checkpoint(self) -> Callable:
         if not self.model_properties.tokenizer:
-            return open_clip.get_tokenizer(self.model_properties.name)
+            # Replace '/'with '-' to support old clip model name style
+            return open_clip.get_tokenizer(self.model_properties.name.replace("/", "-"))
         else:
             logger.info(f"Custom HFTokenizer is provided. Loading...")
             return HFTokenizer(self.model_properties.tokenizer)

diff --git a/tests/core/inference/test_open_clip_model_load.py b/tests/core/inference/test_open_clip_model_load.py
@@ -276,4 +276,15 @@ def test_load_OpenCLIPModel_with_auth_hf(self):
             mock_download_model.assert_called_once_with(
                 repo_location=ModelLocation(**model_properties["model_location"]),
                 auth=model_auth,
-            )
+            )
+
+    def test_load_legacy_openai_clip_model(self):
+        """A test to ensure old OpenAI CLIP models (e.g., ViT-B/32) are loaded correctly."""
+        model_properties = {
+            "name": "ViT-B/32", # Old OpenAI CLIP model name
+            "type": "open_clip",
+            "url": "https://github.com/mlfoundations/open_clip/releases/download/v0.2-weights/vit_b_32-quickgelu-laion400m_e32-46683a32.pt",
+            "dimensions": 512
+        }
+        model = OPEN_CLIP(model_properties=model_properties, device="cpu")
+        model.load()