Embeddings refactor

embeddings.py

@@ -1,53 +1,97 @@
 import datasets
+from datasets import Features, Value, Array1D
 from transformers import CLIPProcessor, CLIPModel
 import torch
 from PIL import Image
+from tqdm import tqdm
 
-# Load the dataset
-dataset = datasets.load_dataset("metmuseum/openaccess")
+# 1) Load the dataset
+dataset = datasets.load_dataset("metmuseum/openaccess", split="train", streaming=False)
+# If the dataset is huge for your machine, consider streaming=True and writing out shards.
 
-# Initialize the CLIP model and processor
-model = CLIPModel.from_pretrained("openai/clip-vit-base-patch32")
-processor = CLIPProcessor.from_pretrained("openai/clip-vit-base-patch32")
+# 2) Initialize model/processor
+model_name = "openai/clip-vit-base-patch32"
+model = CLIPModel.from_pretrained(model_name)
+processor = CLIPProcessor.from_pretrained(model_name)
 
-# Set the model to use the CPU or Apple's MPS if available
-# Change this if you have a fancier computer (:
+# 3) Device + eval
 device = torch.device("mps" if torch.backends.mps.is_available() else "cpu")
 model.to(device)
+model.eval()
 
-# Map to store embeddings with Object ID as key
-embeddings_map = {}
-
-# Function to create embeddings for a PIL Jpeg image
-def create_embedding(image_pil):
-    try:
-        inputs = processor(images=image_pil, return_tensors="pt", padding=True).to(device)
-        with torch.no_grad():
-            embeddings = model.get_image_features(**inputs)
-        return embeddings
-    except Exception as e:
-        print(f"Error processing image: {e}")
-        return None
-
-# Loop through the dataset and process the images, and add them to a map
-# Optionally, you could add more keys here
-# Or, just add the embeddings to the full dataset
-for item in dataset['train']:
-    object_id = item['Object ID']
-    image_pil = item['jpg']
-    if image_pil:
-        embedding = create_embedding(image_pil)
-        if embedding is not None:
-            embeddings_map[object_id] = embedding.cpu().numpy()
-
-# Convert embeddings map to a new dataset
-# Note: I changed to to [embeddings][0] because the examples seemed to like [embeddings] format better
-# than [[[embeddings]]]
-# ...perhaps that's incorrect?
-embedding_dataset = datasets.Dataset.from_dict({
-    'Object ID': list(embeddings_map.keys()),
-    'Embedding': [embedding.tolist() for embedding in embeddings_map.values()][0]
+# 4) Helper to normalize (L2)
+def l2_normalize(x, dim=-1, eps=1e-12):
+    return x / (x.norm(p=2, dim=dim, keepdim=True) + eps)
+
+# 5) Iterate with batching
+BATCH_SIZE = 32  # tune for your machine
+object_ids_batch, images_batch = [], []
+all_object_ids, all_embeddings = [], []
+
+def flush_batch():
+    if not images_batch:
+        return
+    # Processor expects PIL list
+    inputs = processor(images=images_batch, return_tensors="pt")
+    # Only move tensors to device, not the whole dict-of-PIL
+    pixel_values = inputs["pixel_values"].to(device)
+    with torch.no_grad():
+        feats = model.get_image_features(pixel_values=pixel_values)  # (B, D)
+        feats = l2_normalize(feats, dim=-1)                          # normalize
+        feats = feats.cpu()                                          # keep CPU for HF datasets
+    # Save
+    for oid, vec in zip(object_ids_batch, feats):
+        all_object_ids.append(int(oid))
+        all_embeddings.append(vec.numpy().astype("float32"))  # (D,)
+    # clear
+    object_ids_batch.clear()
+    images_batch.clear()
+
+for item in tqdm(dataset):
+    # Depending on the dataset schema, column names may differ.
+    # Using 'Object ID' and 'jpg' from your example; adjust if needed (e.g., 'image').
+    object_id = item.get("Object ID")
+    image_pil = item.get("jpg")
+
+    if object_id is None or image_pil is None:
+        continue
+
+    # Ensure RGB
+    if isinstance(image_pil, Image.Image):
+        img = image_pil.convert("RGB")
+    else:
+        # If it’s an array/bytes, try to convert to PIL; otherwise skip
+        try:
+            img = Image.fromarray(image_pil).convert("RGB")
+        except Exception:
+            continue
+
+    object_ids_batch.append(object_id)
+    images_batch.append(img)
+
+    if len(images_batch) >= BATCH_SIZE:
+        flush_batch()
+
+# flush any remainder
+flush_batch()
+
+# 6) Build a proper HF dataset with explicit features
+if len(all_embeddings) == 0:
+    raise RuntimeError("No embeddings were produced. Check dataset columns and image availability.")
+
+dim = len(all_embeddings[0])
+features = Features({
+    "Object ID": Value("int32"),
+    "Embedding": Array1D(dim, dtype="float32"),
 })
 
-# Save the new dataset to disk
-embedding_dataset.save_to_disk('metmuseum_embeddings')
+embedding_dataset = datasets.Dataset.from_dict(
+    {
+        "Object ID": all_object_ids,
+        "Embedding": all_embeddings,
+    },
+    features=features,
+)
+
+# 7) Save to disk
+embedding_dataset.save_to_disk("metmuseum_embeddings")