Spaces:
Runtime error
Runtime error
DeDeckerThomas
commited on
Commit
·
b96bd14
1
Parent(s):
31decce
Update HF space
Browse files- app.py +58 -89
- css/style.css +19 -0
- requirements.txt +0 -1
app.py
CHANGED
|
@@ -5,10 +5,8 @@ from pipelines.keyphrase_generation_pipeline import KeyphraseGenerationPipeline
|
|
| 5 |
import orjson
|
| 6 |
|
| 7 |
from annotated_text.util import get_annotated_html
|
| 8 |
-
from st_aggrid import AgGrid, GridOptionsBuilder, GridUpdateMode
|
| 9 |
import re
|
| 10 |
import string
|
| 11 |
-
import numpy as np
|
| 12 |
|
| 13 |
|
| 14 |
@st.cache(allow_output_mutation=True, show_spinner=False)
|
|
@@ -21,28 +19,13 @@ def load_pipeline(chosen_model):
|
|
| 21 |
|
| 22 |
def extract_keyphrases():
|
| 23 |
st.session_state.keyphrases = pipe(st.session_state.input_text)
|
| 24 |
-
st.session_state.
|
| 25 |
-
|
| 26 |
-
|
| 27 |
-
|
| 28 |
-
|
| 29 |
-
|
| 30 |
-
|
| 31 |
-
[
|
| 32 |
-
st.session_state.chosen_model,
|
| 33 |
-
st.session_state.input_text,
|
| 34 |
-
],
|
| 35 |
-
st.session_state.keyphrases,
|
| 36 |
-
)
|
| 37 |
-
)
|
| 38 |
-
],
|
| 39 |
-
columns=["model", "text"]
|
| 40 |
-
+ [str(i) for i in range(len(st.session_state.keyphrases))],
|
| 41 |
-
),
|
| 42 |
-
],
|
| 43 |
-
ignore_index=True,
|
| 44 |
-
axis=0,
|
| 45 |
-
).fillna("")
|
| 46 |
|
| 47 |
|
| 48 |
def get_annotated_text(text, keyphrases):
|
|
@@ -90,51 +73,36 @@ def get_annotated_text(text, keyphrases):
|
|
| 90 |
return result
|
| 91 |
|
| 92 |
|
| 93 |
-
def
|
| 94 |
-
|
| 95 |
-
|
| 96 |
-
|
| 97 |
-
and len(st.session_state.selected_rows) == 0
|
| 98 |
-
):
|
| 99 |
-
text, keyphrases = st.session_state.input_text, st.session_state.keyphrases
|
| 100 |
-
else:
|
| 101 |
-
text, keyphrases = (
|
| 102 |
-
st.session_state.selected_rows["text"].values[0],
|
| 103 |
-
[
|
| 104 |
-
keyphrase
|
| 105 |
-
for keyphrase in st.session_state.selected_rows.loc[
|
| 106 |
-
:,
|
| 107 |
-
st.session_state.selected_rows.columns.difference(
|
| 108 |
-
["model", "text"]
|
| 109 |
-
),
|
| 110 |
-
]
|
| 111 |
-
.astype(str)
|
| 112 |
-
.values.tolist()[0]
|
| 113 |
-
if keyphrase != ""
|
| 114 |
-
],
|
| 115 |
-
)
|
| 116 |
|
| 117 |
-
|
|
|
|
| 118 |
|
| 119 |
-
|
| 120 |
-
|
| 121 |
-
|
| 122 |
-
|
| 123 |
-
|
| 124 |
-
|
| 125 |
-
|
| 126 |
-
|
| 127 |
-
|
| 128 |
-
|
| 129 |
-
|
|
|
|
| 130 |
|
| 131 |
|
| 132 |
if "config" not in st.session_state:
|
| 133 |
with open("config.json", "r") as f:
|
| 134 |
content = f.read()
|
| 135 |
st.session_state.config = orjson.loads(content)
|
| 136 |
-
st.session_state.
|
| 137 |
st.session_state.keyphrases = []
|
|
|
|
|
|
|
| 138 |
|
| 139 |
if "select_rows" not in st.session_state:
|
| 140 |
st.session_state.selected_rows = []
|
|
@@ -177,42 +145,43 @@ Do you want to see some magic 🧙♂️? Try it out yourself! 👇
|
|
| 177 |
|
| 178 |
st.write(description)
|
| 179 |
|
| 180 |
-
with st.form("
|
| 181 |
-
|
| 182 |
-
|
| 183 |
-
|
|
|
|
| 184 |
)
|
| 185 |
-
|
| 186 |
st.markdown(
|
| 187 |
-
f"For more information about the chosen model, please be sure to check
|
| 188 |
)
|
| 189 |
|
| 190 |
-
with st.spinner("Loading pipeline..."):
|
| 191 |
-
pipe = load_pipeline(
|
| 192 |
-
f"{st.session_state.config.get('model_author')}/{st.session_state.chosen_model}"
|
| 193 |
-
)
|
| 194 |
-
|
| 195 |
st.session_state.input_text = st.text_area(
|
| 196 |
-
"✍ Input", st.session_state.config.get("example_text"), height=
|
| 197 |
).replace("\n", " ")
|
| 198 |
|
| 199 |
with st.spinner("Extracting keyphrases..."):
|
| 200 |
-
pressed = st.form_submit_button("Extract"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 201 |
|
| 202 |
-
|
| 203 |
-
|
|
|
|
|
|
|
|
|
|
| 204 |
|
| 205 |
-
if len(st.session_state.
|
| 206 |
-
|
| 207 |
-
|
| 208 |
-
|
| 209 |
-
|
| 210 |
-
|
| 211 |
-
|
| 212 |
-
|
| 213 |
-
data = AgGrid(
|
| 214 |
-
st.session_state.data_frame,
|
| 215 |
-
gridOptions=go,
|
| 216 |
-
update_mode=GridUpdateMode.SELECTION_CHANGED,
|
| 217 |
-
)
|
| 218 |
-
st.session_state.selected_rows = pd.DataFrame(data["selected_rows"])
|
|
|
|
| 5 |
import orjson
|
| 6 |
|
| 7 |
from annotated_text.util import get_annotated_html
|
|
|
|
| 8 |
import re
|
| 9 |
import string
|
|
|
|
| 10 |
|
| 11 |
|
| 12 |
@st.cache(allow_output_mutation=True, show_spinner=False)
|
|
|
|
| 19 |
|
| 20 |
def extract_keyphrases():
|
| 21 |
st.session_state.keyphrases = pipe(st.session_state.input_text)
|
| 22 |
+
st.session_state.history[f"run_{st.session_state.current_run_id}"] = {
|
| 23 |
+
"run_id": st.session_state.current_run_id,
|
| 24 |
+
"model": st.session_state.chosen_model,
|
| 25 |
+
"text": st.session_state.input_text,
|
| 26 |
+
"keyphrases": st.session_state.keyphrases,
|
| 27 |
+
}
|
| 28 |
+
st.session_state.current_run_id += 1
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 29 |
|
| 30 |
|
| 31 |
def get_annotated_text(text, keyphrases):
|
|
|
|
| 73 |
return result
|
| 74 |
|
| 75 |
|
| 76 |
+
def render_output(layout, runs, reverse=False, multi_select=False):
|
| 77 |
+
runs = list(runs.values())[::-1] if reverse else list(runs.values())
|
| 78 |
+
for run in runs:
|
| 79 |
+
layout.markdown(f"**⚙️ Output run {run.get('run_id')}**")
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 80 |
|
| 81 |
+
layout.markdown(f"**Model**: {run.get('model')}")
|
| 82 |
+
result = get_annotated_text(run.get("text"), list(run.get("keyphrases")))
|
| 83 |
|
| 84 |
+
layout.markdown(
|
| 85 |
+
get_annotated_html(*result),
|
| 86 |
+
unsafe_allow_html=True,
|
| 87 |
+
)
|
| 88 |
+
if "generation" in st.session_state.chosen_model:
|
| 89 |
+
abstractive_keyphrases = [
|
| 90 |
+
keyphrase
|
| 91 |
+
for keyphrase in run.get("keyphrases")
|
| 92 |
+
if keyphrase.lower() not in run.get("text").lower()
|
| 93 |
+
]
|
| 94 |
+
layout.write(", ".join(abstractive_keyphrases))
|
| 95 |
+
layout.markdown("---")
|
| 96 |
|
| 97 |
|
| 98 |
if "config" not in st.session_state:
|
| 99 |
with open("config.json", "r") as f:
|
| 100 |
content = f.read()
|
| 101 |
st.session_state.config = orjson.loads(content)
|
| 102 |
+
st.session_state.history = {}
|
| 103 |
st.session_state.keyphrases = []
|
| 104 |
+
st.session_state.current_run_id = 1
|
| 105 |
+
st.session_state.chosen_model = st.session_state.config.get("models")[0]
|
| 106 |
|
| 107 |
if "select_rows" not in st.session_state:
|
| 108 |
st.session_state.selected_rows = []
|
|
|
|
| 145 |
|
| 146 |
st.write(description)
|
| 147 |
|
| 148 |
+
with st.form("keyphrase-extraction-form"):
|
| 149 |
+
selectbox_container, _ = st.columns(2)
|
| 150 |
+
|
| 151 |
+
st.session_state.chosen_model = selectbox_container.selectbox(
|
| 152 |
+
"Choose your model:", st.session_state.config.get("models")
|
| 153 |
)
|
| 154 |
+
|
| 155 |
st.markdown(
|
| 156 |
+
f"For more information about the chosen model, please be sure to check out the [🤗 Model Card](https://huggingface.co/DeDeckerThomas/{st.session_state.chosen_model})."
|
| 157 |
)
|
| 158 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 159 |
st.session_state.input_text = st.text_area(
|
| 160 |
+
"✍ Input", st.session_state.config.get("example_text"), height=250
|
| 161 |
).replace("\n", " ")
|
| 162 |
|
| 163 |
with st.spinner("Extracting keyphrases..."):
|
| 164 |
+
pressed = st.form_submit_button("Extract")
|
| 165 |
+
|
| 166 |
+
if pressed:
|
| 167 |
+
with st.spinner("Loading pipeline..."):
|
| 168 |
+
pipe = load_pipeline(
|
| 169 |
+
f"{st.session_state.config.get('model_author')}/{st.session_state.chosen_model}"
|
| 170 |
+
)
|
| 171 |
+
with st.spinner("Extracting keyphrases"):
|
| 172 |
+
extract_keyphrases()
|
| 173 |
|
| 174 |
+
options = st.multiselect(
|
| 175 |
+
"Specify runs you want to see",
|
| 176 |
+
st.session_state.history.keys(),
|
| 177 |
+
format_func=lambda run_id: f"Run {run_id.split('_')[1]}",
|
| 178 |
+
)
|
| 179 |
|
| 180 |
+
if len(st.session_state.history.keys()) > 0:
|
| 181 |
+
if options:
|
| 182 |
+
render_output(
|
| 183 |
+
st,
|
| 184 |
+
{key: st.session_state.history[key] for key in options},
|
| 185 |
+
)
|
| 186 |
+
else:
|
| 187 |
+
render_output(st, st.session_state.history, reverse=True)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
css/style.css
CHANGED
|
@@ -1,5 +1,24 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
@import url('https://fonts.googleapis.com/css2?family=Roboto:ital,wght@0,100;0,300;0,400;0,500;0,700;0,900;1,100;1,300;1,400;1,500;1,700;1,900&display=swap');
|
| 2 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 3 |
body {
|
| 4 |
font-family: 'Roboto', 'Source Sans Pro', sans-serif;
|
| 5 |
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
/*
|
| 2 |
+
|
| 3 |
+
Fonts
|
| 4 |
+
|
| 5 |
+
*/
|
| 6 |
@import url('https://fonts.googleapis.com/css2?family=Roboto:ital,wght@0,100;0,300;0,400;0,500;0,700;0,900;1,100;1,300;1,400;1,500;1,700;1,900&display=swap');
|
| 7 |
|
| 8 |
+
/*
|
| 9 |
+
|
| 10 |
+
HTML body
|
| 11 |
+
|
| 12 |
+
*/
|
| 13 |
body {
|
| 14 |
font-family: 'Roboto', 'Source Sans Pro', sans-serif;
|
| 15 |
}
|
| 16 |
+
|
| 17 |
+
/*
|
| 18 |
+
|
| 19 |
+
Component: Extract Button
|
| 20 |
+
|
| 21 |
+
*/
|
| 22 |
+
.css-1cpxqw2{
|
| 23 |
+
float: right;
|
| 24 |
+
}
|
requirements.txt
CHANGED
|
@@ -3,4 +3,3 @@ transformers[torch]==4.17.0
|
|
| 3 |
pandas==1.4.1
|
| 4 |
numpy==1.22.3
|
| 5 |
st-annotated-text==3.0.0
|
| 6 |
-
streamlit-aggrid==0.2.3.post2
|
|
|
|
| 3 |
pandas==1.4.1
|
| 4 |
numpy==1.22.3
|
| 5 |
st-annotated-text==3.0.0
|
|
|