Upload with huggingface_hub

.gitattributes

@@ -32,3 +32,5 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+cc3m_imagebind_files.json filter=lfs diff=lfs merge=lfs -text
+knn.index filter=lfs diff=lfs merge=lfs -text

build_cc3m_index.py

@@ -0,0 +1,34 @@
+import glob
+import os
+import json
+import tqdm
+import numpy as np
+from multiprocessing import Pool
+from autofaiss import build_index
+
+def load_file(file):
+    return np.load(file)
+
+
+in_dir = '/data0/ImageBindFeatures/cc3m/'
+save_path = "cc3m_imagebind.npy"
+
+files = glob.glob(in_dir + '*/*.npy')
+
+pool = Pool(256)
+results = pool.map(load_file, files)
+results = np.stack(results, axis=0)
+
+np.save(save_path, results)
+
+with open('cc3m_imagebind_files.json', 'w') as f:
+    json.dump([x.replace(in_dir, '') for x in files], f)
+
+
+# build index
+build_index(embeddings="./",
+            index_path="knn.index",
+            index_infos_path="infos.json",
+            max_index_memory_usage = "32G",
+            current_memory_available = "100G",
+            metric_type='ip')

cc3m_imagebind_files.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ed9ea276a776b47c47868fc13bab2c1813f167c3155b6aab58d37dc2939e8df8
+size 129571307

faiss_cc3m_search.py

@@ -0,0 +1,12 @@
+import faiss
+import torch
+
+my_index = faiss.read_index("knn.index")
+
+
+query = torch.rand(3, 1024)
+
+k = 5
+distances, indices = my_index.search(query, k)
+
+import pdb;pdb.set_trace()

infos.json

@@ -0,0 +1 @@
+{"index_key": "HNSW32", "index_param": "efSearch=16", "index_path": "/data1/llama_adapter_project/extract_ImageBind_features/autofaiss_cc3m/autofaiss/knn.index", "size in bytes": 13264998158, "avg_search_speed_ms": 36.55575554355653, "99p_search_speed_ms": 46.38748399913309, "reconstruction error %": 0.0, "nb vectors": 3036761, "vectors dimension": 1024, "compression ratio": 0.9376988151708419}

knn.index

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c39db2288566cebf9aca85306567ce54dd5d2459df8f568717be726af376560b
+size 13264998158