Spaces:
Runtime error
Runtime error
Cache model and tokenizer and lock dependencies
Browse files- app.py +19 -3
- requirements.txt +7 -7
app.py
CHANGED
@@ -2,10 +2,10 @@ import os
|
|
2 |
import sys
|
3 |
|
4 |
import streamlit as st
|
|
|
5 |
from huggingface_hub import snapshot_download
|
6 |
from transformers import AutoTokenizer
|
7 |
|
8 |
-
|
9 |
LOCAL_PATH = snapshot_download("flax-community/clip-spanish")
|
10 |
sys.path.append(LOCAL_PATH)
|
11 |
|
@@ -15,16 +15,24 @@ from test_on_image import run_inference
|
|
15 |
|
16 |
def save_file_to_disk(uplaoded_file):
|
17 |
temp_file = os.path.join("/tmp", uplaoded_file.name)
|
18 |
-
with open(temp_file,"wb") as f:
|
19 |
f.write(uploaded_file.getbuffer())
|
20 |
return temp_file
|
21 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
22 |
def load_tokenizer_and_model():
|
23 |
# load the saved model
|
24 |
tokenizer = AutoTokenizer.from_pretrained("dccuchile/bert-base-spanish-wwm-cased")
|
25 |
model = FlaxHybridCLIP.from_pretrained(LOCAL_PATH)
|
26 |
return tokenizer, model
|
27 |
|
|
|
28 |
tokenizer, model = load_tokenizer_and_model()
|
29 |
|
30 |
st.title("Image-Caption Matching")
|
@@ -36,7 +44,15 @@ if uploaded_file is not None and text_input:
|
|
36 |
try:
|
37 |
local_image_path = save_file_to_disk(uploaded_file)
|
38 |
score = run_inference(local_image_path, text_input, model, tokenizer).tolist()
|
39 |
-
st.image(
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
40 |
st.write(f"## Score: {score:.2f}")
|
41 |
finally:
|
42 |
if local_image_path:
|
|
|
2 |
import sys
|
3 |
|
4 |
import streamlit as st
|
5 |
+
import transformers
|
6 |
from huggingface_hub import snapshot_download
|
7 |
from transformers import AutoTokenizer
|
8 |
|
|
|
9 |
LOCAL_PATH = snapshot_download("flax-community/clip-spanish")
|
10 |
sys.path.append(LOCAL_PATH)
|
11 |
|
|
|
15 |
|
16 |
def save_file_to_disk(uplaoded_file):
|
17 |
temp_file = os.path.join("/tmp", uplaoded_file.name)
|
18 |
+
with open(temp_file, "wb") as f:
|
19 |
f.write(uploaded_file.getbuffer())
|
20 |
return temp_file
|
21 |
|
22 |
+
|
23 |
+
@st.cache(
|
24 |
+
hash_funcs={
|
25 |
+
transformers.models.bert.tokenization_bert_fast.BertTokenizerFast: id,
|
26 |
+
FlaxHybridCLIP: id,
|
27 |
+
}
|
28 |
+
)
|
29 |
def load_tokenizer_and_model():
|
30 |
# load the saved model
|
31 |
tokenizer = AutoTokenizer.from_pretrained("dccuchile/bert-base-spanish-wwm-cased")
|
32 |
model = FlaxHybridCLIP.from_pretrained(LOCAL_PATH)
|
33 |
return tokenizer, model
|
34 |
|
35 |
+
|
36 |
tokenizer, model = load_tokenizer_and_model()
|
37 |
|
38 |
st.title("Image-Caption Matching")
|
|
|
44 |
try:
|
45 |
local_image_path = save_file_to_disk(uploaded_file)
|
46 |
score = run_inference(local_image_path, text_input, model, tokenizer).tolist()
|
47 |
+
st.image(
|
48 |
+
uploaded_file,
|
49 |
+
caption=text_input,
|
50 |
+
width=None,
|
51 |
+
use_column_width=None,
|
52 |
+
clamp=False,
|
53 |
+
channels="RGB",
|
54 |
+
output_format="auto",
|
55 |
+
)
|
56 |
st.write(f"## Score: {score:.2f}")
|
57 |
finally:
|
58 |
if local_image_path:
|
requirements.txt
CHANGED
@@ -1,8 +1,8 @@
|
|
1 |
-
flax
|
2 |
-
|
3 |
-
jax
|
4 |
streamlit==0.84.1
|
5 |
-
torch
|
6 |
-
torchvision
|
7 |
-
transformers
|
8 |
-
watchdog
|
|
|
1 |
+
flax==0.3.4
|
2 |
+
huggingface-hub==0.0.12
|
3 |
+
jax==0.2.17
|
4 |
streamlit==0.84.1
|
5 |
+
torch==1.9.0
|
6 |
+
torchvision==0.10.0
|
7 |
+
transformers==4.8.2
|
8 |
+
watchdog==2.1.3
|