Spaces:
Running
Running
dbleek
commited on
Commit
•
0d30c2b
1
Parent(s):
1e95f51
added new classifier
Browse files- milestone-3.py +25 -22
- patent_classifier_v4.pt +3 -0
milestone-3.py
CHANGED
@@ -1,8 +1,8 @@
|
|
1 |
import streamlit as st
|
2 |
import torch
|
|
|
3 |
from datasets import load_dataset
|
4 |
from transformers import AutoTokenizer
|
5 |
-
from transformers import AutoModelForSequenceClassification
|
6 |
from transformers import pipeline
|
7 |
|
8 |
# Load HUPD dataset
|
@@ -21,21 +21,24 @@ dataset_dict = load_dataset(
|
|
21 |
filtered_dataset = dataset_dict["validation"].filter(
|
22 |
lambda e: e["decision"] == "ACCEPTED" or e["decision"] == "REJECTED"
|
23 |
)
|
24 |
-
|
|
|
|
|
|
|
25 |
dataset = dataset.sort("patent_number")
|
26 |
|
27 |
# Create pipeline using model trainned on Colab
|
28 |
-
model = torch.load("
|
29 |
tokenizer = AutoTokenizer.from_pretrained("distilbert-base-uncased")
|
30 |
-
|
|
|
31 |
|
32 |
-
|
33 |
-
def load_patent():
|
34 |
selected_application = dataset.select([applications[st.session_state.id]])
|
35 |
st.session_state.abstract = selected_application["abstract"][0]
|
36 |
st.session_state.claims = selected_application["claims"][0]
|
37 |
st.session_state.title = selected_application["title"][0]
|
38 |
-
|
39 |
|
40 |
st.title("CS-GY-6613 Project Milestone 3")
|
41 |
|
@@ -44,11 +47,12 @@ applications = {}
|
|
44 |
for ds_index, example in enumerate(dataset):
|
45 |
applications.update({example["patent_number"]: ds_index})
|
46 |
st.selectbox(
|
47 |
-
"Select a patent application:", applications, on_change=
|
48 |
)
|
49 |
|
50 |
-
#
|
51 |
-
st.
|
|
|
52 |
|
53 |
# Classifier input form
|
54 |
with st.form("Input Form"):
|
@@ -61,16 +65,15 @@ with st.form("Input Form"):
|
|
61 |
submitted = st.form_submit_button("Get Patentability Score")
|
62 |
|
63 |
if submitted:
|
64 |
-
|
65 |
-
|
66 |
-
|
67 |
-
|
68 |
-
|
69 |
-
|
70 |
-
|
71 |
-
|
72 |
-
result = st.markdown(
|
73 |
-
"This text was classified as **{}** with a confidence score of **{}**.".format(
|
74 |
-
pred, score
|
75 |
-
)
|
76 |
)
|
|
|
|
|
|
|
|
1 |
import streamlit as st
|
2 |
import torch
|
3 |
+
from datasets import combine
|
4 |
from datasets import load_dataset
|
5 |
from transformers import AutoTokenizer
|
|
|
6 |
from transformers import pipeline
|
7 |
|
8 |
# Load HUPD dataset
|
|
|
21 |
filtered_dataset = dataset_dict["validation"].filter(
|
22 |
lambda e: e["decision"] == "ACCEPTED" or e["decision"] == "REJECTED"
|
23 |
)
|
24 |
+
seed = 88
|
25 |
+
accepted = filtered_dataset.filter(lambda e: e["decision"] == "ACCEPTED").shuffle(seed).select(range(5))
|
26 |
+
rejected = filtered_dataset.filter(lambda e: e["decision"] == "REJECTED").shuffle(seed).select(range(5))
|
27 |
+
dataset = combine.concatenate_datasets([accepted, rejected])
|
28 |
dataset = dataset.sort("patent_number")
|
29 |
|
30 |
# Create pipeline using model trainned on Colab
|
31 |
+
model = torch.load("patent_classifier_v4.pt", map_location=torch.device("cpu"))
|
32 |
tokenizer = AutoTokenizer.from_pretrained("distilbert-base-uncased")
|
33 |
+
tokenizer_kwargs = {'padding':True,'truncation':True}
|
34 |
+
classifier = pipeline("text-classification", model=model, tokenizer=tokenizer, **tokenizer_kwargs)
|
35 |
|
36 |
+
def load_data():
|
|
|
37 |
selected_application = dataset.select([applications[st.session_state.id]])
|
38 |
st.session_state.abstract = selected_application["abstract"][0]
|
39 |
st.session_state.claims = selected_application["claims"][0]
|
40 |
st.session_state.title = selected_application["title"][0]
|
41 |
+
st.session_state.decision = selected_application["decision"][0]
|
42 |
|
43 |
st.title("CS-GY-6613 Project Milestone 3")
|
44 |
|
|
|
47 |
for ds_index, example in enumerate(dataset):
|
48 |
applications.update({example["patent_number"]: ds_index})
|
49 |
st.selectbox(
|
50 |
+
"Select a sample patent application:", applications, on_change=load_data, key="id"
|
51 |
)
|
52 |
|
53 |
+
# Sample title/decision displayed for additional context only, not used with model
|
54 |
+
st.text_input("Sample Title", key="title", value=dataset[0]["title"])
|
55 |
+
st.text_input("Sample Decision", key="decision", value=dataset[0]["decision"])
|
56 |
|
57 |
# Classifier input form
|
58 |
with st.form("Input Form"):
|
|
|
65 |
submitted = st.form_submit_button("Get Patentability Score")
|
66 |
|
67 |
if submitted:
|
68 |
+
tokens = tokenizer(abstract, claims, return_tensors='pt', **tokenizer_kwargs)
|
69 |
+
with torch.no_grad():
|
70 |
+
output = model(**tokens)
|
71 |
+
logits = output.logits
|
72 |
+
pred = torch.softmax(logits, dim=1)
|
73 |
+
score = pred[0][1] # index 1 of softmax output is probability that decision = ACCEPTED
|
74 |
+
st.markdown(
|
75 |
+
"This application's patentability score is **{}**".format(score)
|
|
|
|
|
|
|
|
|
76 |
)
|
77 |
+
|
78 |
+
|
79 |
+
|
patent_classifier_v4.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:ae0d471894ba6a7847254acda873e574837547b684b854eaa96efe3b593f8c2d
|
3 |
+
size 267882526
|