Spaces:
Runtime error
Runtime error
update paths
Browse files
app.py
CHANGED
@@ -4,40 +4,50 @@ import numpy as np
|
|
4 |
import h5py
|
5 |
import faiss
|
6 |
from PIL import Image
|
7 |
-
import io
|
8 |
import pickle
|
9 |
import random
|
10 |
|
|
|
11 |
def getRandID():
|
12 |
indx = random.randrange(0, 396503)
|
13 |
return indx_to_id_dict[indx], indx
|
14 |
|
|
|
15 |
def chooseImageIndex(indexType):
|
16 |
-
if
|
17 |
return image_index_IP
|
18 |
-
elif
|
|
|
19 |
return image_index_L2
|
20 |
-
elif
|
|
|
21 |
return image_index_HNSW
|
22 |
-
elif
|
|
|
23 |
return image_index_IVF
|
24 |
-
elif
|
|
|
25 |
return image_index_LSH
|
26 |
|
|
|
27 |
def chooseDNAIndex(indexType):
|
28 |
-
if
|
29 |
return dna_index_IP
|
30 |
-
elif
|
|
|
31 |
return dna_index_L2
|
32 |
-
elif
|
|
|
33 |
return dna_index_HNSW
|
34 |
-
elif
|
|
|
35 |
return dna_index_IVF
|
36 |
-
elif
|
|
|
37 |
return dna_index_LSH
|
38 |
|
39 |
|
40 |
-
|
41 |
def searchEmbeddings(id, mod1, mod2, indexType):
|
42 |
# variable and index initialization
|
43 |
dim = 768
|
@@ -47,16 +57,15 @@ def searchEmbeddings(id, mod1, mod2, indexType):
|
|
47 |
index = faiss.IndexFlatIP(dim)
|
48 |
|
49 |
# get index
|
50 |
-
if
|
51 |
index = chooseImageIndex(indexType)
|
52 |
-
elif
|
53 |
index = chooseDNAIndex(indexType)
|
54 |
-
|
55 |
|
56 |
# search for query
|
57 |
-
if
|
58 |
query = id_to_image_emb_dict[id]
|
59 |
-
elif
|
60 |
query = id_to_dna_emb_dict[id]
|
61 |
query = query.astype(np.float32)
|
62 |
D, I = index.search(query, num_neighbors)
|
@@ -66,25 +75,26 @@ def searchEmbeddings(id, mod1, mod2, indexType):
|
|
66 |
for indx in I[0]:
|
67 |
id = indx_to_id_dict[indx]
|
68 |
id_list.append(id)
|
69 |
-
|
70 |
return id_list
|
71 |
|
|
|
72 |
with gr.Blocks() as demo:
|
73 |
|
74 |
# for hf: change all file paths, indx_to_id_dict as well
|
75 |
|
76 |
# load indexes
|
77 |
-
image_index_IP = faiss.read_index("
|
78 |
-
image_index_L2 = faiss.read_index("big_image_index_FlatL2.index")
|
79 |
-
image_index_HNSW = faiss.read_index("big_image_index_HNSWFlat.index")
|
80 |
-
image_index_IVF = faiss.read_index("big_image_index_IVFFlat.index")
|
81 |
-
image_index_LSH = faiss.read_index("big_image_index_LSH.index")
|
82 |
-
|
83 |
-
dna_index_IP = faiss.read_index("
|
84 |
-
dna_index_L2 = faiss.read_index("big_dna_index_FlatL2.index")
|
85 |
-
dna_index_HNSW = faiss.read_index("big_dna_index_HNSWFlat.index")
|
86 |
-
dna_index_IVF = faiss.read_index("big_dna_index_IVFFlat.index")
|
87 |
-
dna_index_LSH = faiss.read_index("big_dna_index_LSH.index")
|
88 |
|
89 |
with open("dataset_processid_list.pickle", "rb") as f:
|
90 |
dataset_processid_list = pickle.load(f)
|
@@ -109,14 +119,15 @@ with gr.Blocks() as demo:
|
|
109 |
mod1 = gr.Radio(choices=["DNA", "Image"], label="Search From:")
|
110 |
mod2 = gr.Radio(choices=["DNA", "Image"], label="Search To:")
|
111 |
|
112 |
-
indexType = gr.Radio(
|
|
|
|
|
113 |
process_id = gr.Textbox(label="ID:", info="Enter a sample ID to search for")
|
114 |
-
process_id_list = gr.Textbox(label="Closest 10 matches:"
|
115 |
-
search_btn = gr.Button("Search")
|
116 |
id_btn.click(fn=getRandID, inputs=[], outputs=[rand_id, rand_id_indx])
|
117 |
|
118 |
-
search_btn.click(fn=searchEmbeddings, inputs=[process_id, mod1, mod2, indexType],
|
119 |
-
|
120 |
-
|
121 |
|
122 |
-
demo.launch()
|
|
|
4 |
import h5py
|
5 |
import faiss
|
6 |
from PIL import Image
|
7 |
+
import io
|
8 |
import pickle
|
9 |
import random
|
10 |
|
11 |
+
|
12 |
def getRandID():
|
13 |
indx = random.randrange(0, 396503)
|
14 |
return indx_to_id_dict[indx], indx
|
15 |
|
16 |
+
|
17 |
def chooseImageIndex(indexType):
|
18 |
+
if indexType == "FlatIP(default)":
|
19 |
return image_index_IP
|
20 |
+
elif indexType == "FlatL2":
|
21 |
+
raise NotImplementedError
|
22 |
return image_index_L2
|
23 |
+
elif indexType == "HNSWFlat":
|
24 |
+
raise NotImplementedError
|
25 |
return image_index_HNSW
|
26 |
+
elif indexType == "IVFFlat":
|
27 |
+
raise NotImplementedError
|
28 |
return image_index_IVF
|
29 |
+
elif indexType == "LSH":
|
30 |
+
raise NotImplementedError
|
31 |
return image_index_LSH
|
32 |
|
33 |
+
|
34 |
def chooseDNAIndex(indexType):
|
35 |
+
if indexType == "FlatIP(default)":
|
36 |
return dna_index_IP
|
37 |
+
elif indexType == "FlatL2":
|
38 |
+
raise NotImplementedError
|
39 |
return dna_index_L2
|
40 |
+
elif indexType == "HNSWFlat":
|
41 |
+
raise NotImplementedError
|
42 |
return dna_index_HNSW
|
43 |
+
elif indexType == "IVFFlat":
|
44 |
+
raise NotImplementedError
|
45 |
return dna_index_IVF
|
46 |
+
elif indexType == "LSH":
|
47 |
+
raise NotImplementedError
|
48 |
return dna_index_LSH
|
49 |
|
50 |
|
|
|
51 |
def searchEmbeddings(id, mod1, mod2, indexType):
|
52 |
# variable and index initialization
|
53 |
dim = 768
|
|
|
57 |
index = faiss.IndexFlatIP(dim)
|
58 |
|
59 |
# get index
|
60 |
+
if mod2 == "Image":
|
61 |
index = chooseImageIndex(indexType)
|
62 |
+
elif mod2 == "DNA":
|
63 |
index = chooseDNAIndex(indexType)
|
|
|
64 |
|
65 |
# search for query
|
66 |
+
if mod1 == "Image":
|
67 |
query = id_to_image_emb_dict[id]
|
68 |
+
elif mod1 == "DNA":
|
69 |
query = id_to_dna_emb_dict[id]
|
70 |
query = query.astype(np.float32)
|
71 |
D, I = index.search(query, num_neighbors)
|
|
|
75 |
for indx in I[0]:
|
76 |
id = indx_to_id_dict[indx]
|
77 |
id_list.append(id)
|
78 |
+
|
79 |
return id_list
|
80 |
|
81 |
+
|
82 |
with gr.Blocks() as demo:
|
83 |
|
84 |
# for hf: change all file paths, indx_to_id_dict as well
|
85 |
|
86 |
# load indexes
|
87 |
+
image_index_IP = faiss.read_index("bioscan_5m_image_IndexFlatIP.index")
|
88 |
+
# image_index_L2 = faiss.read_index("big_image_index_FlatL2.index")
|
89 |
+
# image_index_HNSW = faiss.read_index("big_image_index_HNSWFlat.index")
|
90 |
+
# image_index_IVF = faiss.read_index("big_image_index_IVFFlat.index")
|
91 |
+
# image_index_LSH = faiss.read_index("big_image_index_LSH.index")
|
92 |
+
|
93 |
+
dna_index_IP = faiss.read_index("bioscan_5m_dna_IndexFlatIP.index")
|
94 |
+
# dna_index_L2 = faiss.read_index("big_dna_index_FlatL2.index")
|
95 |
+
# dna_index_HNSW = faiss.read_index("big_dna_index_HNSWFlat.index")
|
96 |
+
# dna_index_IVF = faiss.read_index("big_dna_index_IVFFlat.index")
|
97 |
+
# dna_index_LSH = faiss.read_index("big_dna_index_LSH.index")
|
98 |
|
99 |
with open("dataset_processid_list.pickle", "rb") as f:
|
100 |
dataset_processid_list = pickle.load(f)
|
|
|
119 |
mod1 = gr.Radio(choices=["DNA", "Image"], label="Search From:")
|
120 |
mod2 = gr.Radio(choices=["DNA", "Image"], label="Search To:")
|
121 |
|
122 |
+
indexType = gr.Radio(
|
123 |
+
choices=["FlatIP(default)", "FlatL2", "HNSWFlat", "IVFFlat", "LSH"], label="Index:", value="FlatIP(default)"
|
124 |
+
)
|
125 |
process_id = gr.Textbox(label="ID:", info="Enter a sample ID to search for")
|
126 |
+
process_id_list = gr.Textbox(label="Closest 10 matches:")
|
127 |
+
search_btn = gr.Button("Search")
|
128 |
id_btn.click(fn=getRandID, inputs=[], outputs=[rand_id, rand_id_indx])
|
129 |
|
130 |
+
search_btn.click(fn=searchEmbeddings, inputs=[process_id, mod1, mod2, indexType], outputs=[process_id_list])
|
131 |
+
|
|
|
132 |
|
133 |
+
demo.launch()
|